The public record of rogue AI activity.

Georgia deputies allegedly misused license plate reader data for non-law-enforcement purposes, audit finds

AI loose in the world·Data exfiltrationDERIVED

An internal audit by the Cherokee County Sheriff's Office in Georgia reportedly found that three deputies used the agency's automated license plate reader database for non-law-enforcement purposes, in violation of policy and state law. The incident date is listed as 2026-06-12.

What this means for youCameras that read license plates feed databases that staff can search, and people with access can misuse them. Local governments can ask who may search such data and how often searches are audited.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

15 malicious JetBrains plugins silently exfiltrated developers' OpenAI, DeepSeek and SiliconFlow API keys

AI loose in the world·Data exfiltrationDERIVED

Security firm Aikido Security discovered 15 malicious plugins on the JetBrains Marketplace that posed as AI coding assistants offering chat, commit messages, code review, bug finding and unit tests. Each functioned as advertised, but when a developer entered an AI provider API key into the plugin settings, it was silently forwarded over plain HTTP to an attacker-controlled server. Stolen keys covered OpenAI, DeepSeek and SiliconFlow. The two most-downloaded plugins, DeepSeek AI Assist and CodeGPT AI Assistant, reported over 25,000 downloads each, though Aikido noted counts can be inflated. The campaign ran from late October 2025 through June 2026.

What this means for youIf you use AI coding assistants in JetBrains IDEs, uninstall unfamiliar plugins and rotate any OpenAI, DeepSeek, or SiliconFlow keys you entered into them. At work, teams should audit plugin installs and restrict where API keys are stored.

Source: BleepingComputer

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Attacker reportedly used LLM agent to steal internal database after breaching marimo Python notebook

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

An unidentified attacker reportedly used an LLM agent in a May 10 intrusion that began with a vulnerable marimo Python notebook. The agent reportedly reused AWS credentials to get an SSH key, pivoted through a bastion host and exfiltrated an internal PostgreSQL database. Sysdig attributed the commands to real-time agent execution.

What this means for youAttackers are reportedly using AI agents to move through networks on their own, step by step. Small businesses should patch exposed tools quickly, and check that cloud credentials are limited in scope and can be rotated.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Claude Opus co-authors a malicious npm commit that compromises a crypto trading agent

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Anthropic's Claude Opus AI model co-authored a code commit that introduced a malicious npm package dependency into an autonomous crypto trading agent. The malware, linked to North Korean group Famous Chollima, enabled theft of crypto assets and sensitive data -- a direct harm to property caused by AI-assisted software development where the AI model made a malicious package a dependency, giving attackers access to users' crypto wallets and funds.

What this means for youAt home: if you use AI-built crypto or trading tools, a hidden bad code package could drain your wallet. At work: review every dependency an AI assistant adds before it ships. In government: this shows supply-chain risk from AI-written code. Check that any AI-written code has a human review of new packages, and keep wallet funds in accounts the agent cannot reach.

Source: OECD.AI Incidents Monitor (corroborated by Infosecurity Magazine and Cryptopolitan)

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Vercel breach started in a third-party AI tool and ended in customers' environment variables

AI loose in the world·Data exfiltrationDERIVED

Vercel disclosed that a security incident originated with a compromise of Context.ai, a third-party AI tool used by a Vercel employee. The attacker used that access to take over the employee's individual Google Workspace account, then their Vercel account, pivoted into a Vercel environment, and enumerated and decrypted non-sensitive environment variables belonging to a subset of customers, who were told to rotate credentials immediately. Vercel engaged Google Mandiant, notified law enforcement, and published IOCs; its investigation later identified a small number of additional compromised accounts.

What this means for youIf you use Vercel or a similar hosting service, rotate any secrets stored in its environment settings, especially if you connected a third-party AI tool to your accounts. At work, check which outside AI tools have been granted access to company email, cloud, or code accounts.

Source: Vercel security bulletin (Knowledge Base)

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Compromised LiteLLM PyPI packages harvest credentials via supply-chain attack

AI loose in the world·Data exfiltrationDERIVED

Malicious versions of the litellm Python package (1.82.7 and 1.82.8) were published to PyPI using compromised maintainer credentials, part of the TeamPCP campaign. The payload harvested SSH keys, AWS/GCP/Azure credentials, Kubernetes tokens and database passwords, then exfiltrated them to an attacker-controlled domain. The packages were live for roughly 40 minutes before PyPI quarantined them.

What this means for youAt work, if a developer or team installed litellm 1.82.7 or 1.82.8, treat every credential on that machine as exposed and rotate them. At home, this matters only if you run LiteLLM yourself; check your installed version and rotate SSH, cloud, and database passwords if it matches.

Source: LiteLLM (security update)

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Meta AI Smart Glasses Reportedly Routed Intimate Imagery to Reviewers at Kenyan Contractor Sama Before Meta Ended Contract

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Meta AI smart glasses reportedly sent media and transcripts from AI interactions to Sama contractors in Kenya for human review. Workers said they saw nudity, bathroom use, sex, bank cards, and other private activity, sometimes involving people unaware they were recorded or faces left visible despite filtering. The reports prompted lawsuits and regulatory inquiries before Meta paused the work and ended its Sama contract, affecting 1,108 workers.

What this means for youIf you wear AI smart glasses, private moments you record may be sent to human reviewers, including people in other countries. At work, employers should check whether such devices can capture client or staff data. Governments may need to ask companies where recordings go and who can see them.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Grok Reportedly Disclosed Adult Performer Siri Dahl's Legal Name and Birthdate, Allegedly Contributing to Doxxing and Harassment

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Grok is reported to have publicly provided adult performer Siri Dahl's legal name and birthdate without being asked for that information. Dahl reportedly said she had worked to keep those details private and that, after the disclosure, impersonation accounts and reposts of stolen content using her legal name purportedly appeared online.

What this means for youAt home: an AI chatbot can reveal personal details about you that you tried to keep private, and others may use them to harass you. At work: staff handling sensitive identities should not assume chatbots protect them. In government: this adds to the case for privacy rules on AI answers. Check your privacy settings and ask platforms to remove personal data from AI outputs.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Moltbook's public database exposed 1.5 million AI-agent API keys in plaintext

AI loose in the world·Data exfiltrationDERIVED

Wiz researchers found that Moltbook, a viral social network where AI agents post and interact, ran a database with public read access and no row-level security. It exposed roughly 1.5 million API keys in plaintext - for OpenAI, Anthropic, AWS, GitHub, and Google Cloud - alongside private agent-to-agent messages and owner email addresses, and the same misconfiguration allowed writes to any post. Anyone who found the endpoint could impersonate arbitrary agents or reach the accounts behind their keys, and the keys stayed valid until each owner rotated them. Access was locked down within about a day of disclosure.

What this means for youIf you connected an AI agent to services like OpenAI, Anthropic, or cloud accounts through Moltbook, those keys may have been exposed. Rotate them and check those accounts for unexpected use or charges.

Source: Wiz

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

NPR Host David Greene Alleged Google's NotebookLM Replicated His Voice Without Consent, Prompting Lawsuit

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

NPR's David Greene reportedly sued Google LLC and Alphabet in Santa Clara County, alleging NotebookLM's Audio Overviews uses a synthetic male voice that purportedly mimics his cadence and delivery without consent or compensation. The complaint reportedly cited an independent voice-recognition analysis reporting 53–60% confidence his voice trained the model. Google reportedly called the allegations baseless and said the voice is based on a paid actor.

What this means for youIf you record your voice for a podcast, job, or app, check whether its terms allow AI training on it before you agree. Courts are now testing whether a person's voice can be copied into AI products without permission.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Border Patrol Agent Allegedly Claimed Facial Recognition Identified Minneapolis ICE Observer and Global Entry Was Reportedly Revoked Three Days Later

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

On 01/10/2026 near Minneapolis, Minnesota, legal observer Nicole Cleland reportedly stated that a CBP Border Patrol agent stopped her vehicle, addressed her by name, and claimed agents used facial recognition to identify her while recording on body cam. She reportedly had her Global Entry and TSA PreCheck revoked on 01/13/2026, which she alleged was retaliatory intimidation, causing travel burden and chilling effects.

What this means for youAt home, this shows police and border agents may use facial recognition to identify people who record them, and a travel status can be affected afterward. At work and in government, it raises questions about how these tools are limited and whether people can challenge them.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Perplexity AI Reportedly Accused in Federal Lawsuit of Purported Copyright Infringement and False Attribution of Chicago Tribune Content

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

The Chicago Tribune filed a federal lawsuit alleging that Perplexity AI unlawfully reproduced and paraphrased its copyrighted journalism in generative chatbot and search outputs. The complaint claims the AI system produced substitutive answers that bypassed links to the Tribune's website, diverted revenue, and at times hallucinated inaccurate information falsely attributed to the newspaper.

What this means for youFor people who get news from AI search tools, answers can reuse publishers' work and sometimes misattribute claims, so check the original article before trusting or sharing an AI summary. For businesses and publishers, this shows how AI answers can divert readers and revenue away from source sites. For government, it is a live test of how copyright law applies to AI answer engines.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Walmart AI voice system accused in suit of collecting Illinois callers' voiceprints without consent

AI loose in the world·Data exfiltrationDERIVED

A proposed class action alleges Walmart's AI voice system captured biometric voiceprints from Illinois residents who called local stores, without meaningful consent or an explanation of retention and deletion. Plaintiff Amber Smith says her voiceprint was captured on two calls in December 2025 and one in July 2026. The claims are allegations.

What this means for youIf you call a store or company, an automated voice system may be recording and analysing your voice. Check whether a business tells you how it uses voice data, and ask how to opt out or have it deleted.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Google's Antigravity IDE vulnerable to data exfiltration and remote code execution via prompt injection

Rogue: Went past its access, Caused harmAI loose in the world·Data exfiltrationDERIVED

Security researchers including Johann Rehberger (Embrace The Red) and PromptArmor demonstrated that Google's Antigravity agentic IDE could be manipulated through indirect prompt injection to exfiltrate sensitive data from a developer's environment. Attackers planted hidden instructions in a seemingly harmless document that coerced the agent into reading the developer's .env file and leaking secrets. Rehberger documented five vulnerability classes including remote code execution, auto-execute bypasses, and data exfiltration via the read_url_content tool and image rendering. Google's own bug-hunters page listed the exfiltration and code-execution paths as known issues under active remediation.

What this means for youAt home and work, an AI coding tool that reads untrusted documents can be tricked into leaking passwords and keys stored in project files. Governments and companies should treat coding agents as software that needs access limits. Check what files and secrets your AI coding tool can read, and keep keys out of project folders.

Source: Embrace The Red

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Secret Desires AI Platform Reportedly Exposed Nearly Two Million Sensitive Images in Cloud Storage Leak

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

The erotic AI chatbot and image-generation platform Secret Desires reportedly left nearly two million sensitive images and videos publicly exposed in misconfigured cloud storage. The leaked files reportedly included personal photos, workplace and university information, and explicit AI-generated deepfakes of women and girls. The content reportedly became inaccessible shortly after journalists contacted the platform.

What this means for youIf you have used AI image or chat apps, the files you uploaded could be exposed by careless storage settings. At work, check whether staff have used such tools with personal or company photos.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Lawsuit alleges Fireflies.AI meeting assistant collected an Illinois participant's voice data without consent

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A lawsuit alleges that on November 18, 2025, Fireflies.AI recorded and analyzed the voice of Illinois resident Katelin Cruz in a virtual meeting hosted by Western Illinois Dreamers. She had no account and gave no written consent. The suit says speaker recognition created and kept biometric voice data.

What this means for youIf you join a meeting where an AI note-taker is running, it may be recording and identifying your voice even if you never signed up. Allegations only; the court has not decided. Ask meeting hosts whether an AI assistant is on, and ask them to turn it off if you have not agreed.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

ChatGPT Reportedly Found to Reproduce Protected German Lyrics in Copyright Case

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A Munich regional court ruled that ChatGPT reportedly reproduced protected German song lyrics and that OpenAI's models were trained on copyrighted texts, including works by musician Herbert Grönemeyer, without authorization. The court reportedly found both memorization of nine songs and lyric output to infringe exploitation rights. OpenAI disputes the ruling and may appeal. Damages were ordered, with implications for AI training on copyrighted works.

What this means for youSong lyrics, books, and articles stay protected when an AI reproduces them, and the ruling could shape how AI tools are trained and sold. At work, check that AI-written text quoting songs or articles is not copied word for word before you publish or reuse it.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

State-sponsored group automates ~80–90% of a cyber-espionage operation using Claude Code and MCP tools

AI loose in the world·Data exfiltrationDERIVED

Anthropic disclosed that a Chinese state-sponsored group abused its Claude Code agent and MCP tooling to automate roughly 80–90% of a multi-stage agentic cyber-espionage operation targeting around 30 global organizations, including tech, finance, chemical and government entities. Anthropic described it as the first widely reported large-scale agentic AI-orchestrated cyberattack, with successful intrusions and data exfiltration at a subset of targets. Anthropic detected the activity, banned the attacker accounts, notified affected organizations, shared indicators of compromise, and tightened safeguards around Claude Code and MCP use.

What this means for youAt home, little changes directly, but data held by banks, tech firms and agencies can be targeted faster by AI-driven attacks. At work, expect attackers to move quicker, so keep software updated and use multi-factor sign-in. In government, this adds pressure for rules and information-sharing on AI misuse.

Source: Anthropic

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Prompt injection in Manus turned its sandbox into a publicly-exposed VS Code server with API keys inside

Rogue: Acted on its own, Went past its accessAI loose in the world·Data exfiltrationDERIVED

Johann Rehberger demonstrated a full kill chain against the Manus agent: an indirect prompt injection in a document Manus was processing caused it to expose its VS Code development server to the public internet, with the sandbox's environment variables — including API keys — readable. The issue was disclosed to Manus on June 1, 2025; the write-up was published on August 25, 2025 during the Month of AI Bugs.

What this means for youAt home and work, AI agents that read documents can be tricked by hidden instructions into exposing your accounts or keys. Businesses should limit what keys and files an agent can reach; governments should expect agent security rules to matter. Check what secrets your AI agent can access and keep them out of its workspace.

Source: Embrace The Red (Johann Rehberger)

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Windsurf Cascade coding agent leaks developer secrets via indirect prompt injection

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Johann Rehberger disclosed that Windsurf Cascade, the coding agent inside the Windsurf editor (a VS Code fork), allowed an adversary to exfiltrate data from the developer's machine through indirect prompt injection. The vulnerabilities follow the 'lethal trifecta' pattern and were responsibly disclosed on May 30, 2025; Rehberger said receipt was acknowledged but subsequent status inquiries went unanswered, amid business disruption and the departure of Windsurf's CEO and core team.

What this means for youIf you or your team use AI coding assistants, they can be tricked by instructions hidden in files or web pages and may send secrets out. Check that the tool's update is installed, limit what it can read and send, and avoid running it on untrusted code or projects with live credentials.

Source: Embrace The Red (Johann Rehberger)

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Google Jules asynchronous coding agent vulnerable to multiple data exfiltration issues

AI loose in the world·Data exfiltrationDERIVED

Johann Rehberger disclosed that Google Jules, Google's asynchronous coding agent, was vulnerable to multiple data exfiltration issues. Jules operates as a main agent that proposes a plan and then hands subtasks off to worker agents carrying different tools, and its system prompt and tooling could be steered via prompt injection to pull data off the developer's machine — an instance of the 'lethal trifecta' of agent risks. Findings were shared with Google before publication.

What this means for youIf you use AI coding agents that can read your project files, a hidden instruction in a web page or code comment could trick them into sending data out. At work, teams using these tools should limit what the agent can access and review what it sends.

Source: Embrace The Red (Johann Rehberger)

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Stolen OAuth tokens for Drift AI chatbot expose 700+ Salesforce customers

AI loose in the world·Data exfiltrationDERIVED

The threat actor tracked as UNC6395 used compromised OAuth and refresh tokens for Salesloft's Drift AI chat agent to impersonate the trusted integration and query the Salesforce instances of more than 700 organizations between August 8 and 18, 2025, exfiltrating stored secrets including AWS keys and Snowflake tokens. On August 20 Salesloft and Salesforce revoked all Drift tokens and pulled the app from the AppExchange.

What this means for youIf your company uses Salesforce or connected AI tools, its stored passwords and access keys may have been copied, so ask whether it has rotated them. At home, this matters mainly if you are a customer of an affected business and it may contact you about exposed data.

Source: Google Threat Intelligence Group

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Microsoft's Windows Recall Allegedly Stores Passwords and Social Security Numbers in Preview Mode

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Microsoft's Windows Recall, an AI-powered screenshot and retrieval tool for Copilot+ PCs, was allegedly still capturing sensitive information such as passwords, Social Security numbers, and bank details despite a built-in "filter sensitive information" feature. Independent testing reportedly found the filter failed in multiple cases. Microsoft reportedly classified Recall as a preview feature and said improvements were in progress.

What this means for youAt home, Recall may keep screenshots of your passwords, bank details and ID numbers on your PC, where others could find them. At work, this could put customer or company data at risk. Check whether Recall is turned on in Windows settings and turn it off if you handle sensitive information.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Gaggle AI Monitoring at Lawrence, Kansas High School Reportedly Misflags Student Content and Blocks Emails

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

In Lawrence, Kansas, students allege the Gaggle Safety Management AI wrongly flagged benign schoolwork, including art photos and casual messages, as child pornography or threats. The system reportedly deleted content, blocked an email records request, and led to questioning of students. Critics cite chilling effects and privacy risks. A lawsuit filed in August 2025 challenges the district's use of Gaggle as unconstitutional surveillance. Gaggle reportedly denies compromising privacy.

What this means for youStudents and families may find schoolwork flagged or deleted by an AI monitor, and a wrong flag can lead to questioning, so the stakes are real for kids. Schools using these tools should be able to explain what is scanned and how records are kept. Ask your school whether a person reviews AI flags before any action is taken, and request a copy of any flagged records and the retention policy for them.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Reported Public Exposure of Over 100,000 LLM Conversations via Share Links Indexed by Search Engines and Archived

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Across 2024 and 2025, the share features in multiple LLM platforms, including ChatGPT, Claude, Copilot, Qwen, Mistral, and Grok, allegedly exposed user conversations marked "discoverable" to search engines and archiving services. Over 100,000 chats were reportedly indexed and later scraped, purportedly revealing API keys, access tokens, personal identifiers, and sensitive business data.

What this means for youAt home, a chat you share by link may become publicly searchable, so avoid putting passwords or personal details in chats. At work, staff pasting keys or client data into AI chats risk leaks, so check sharing settings and company rules.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

CISA Acting Director Reportedly Uploaded Sensitive Government Documents to Public ChatGPT Instance

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Madhu Gottumukkala, acting director of the Cybersecurity and Infrastructure Security Agency (CISA), reportedly uploaded government contracting documents marked "for official use only" into a public version of ChatGPT. The uploads reportedly triggered automated cybersecurity alerts and prompted a Department of Homeland Security review to assess potential exposure of sensitive information.

What this means for youGovernment staff may paste sensitive files into public AI tools without realizing the data can leave official control. At work, check your organization's rules on which documents can go into outside AI services before you upload anything.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

McDonald's McHire chatbot platform exposed 64 million job applications behind the password '123456'

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Researchers Ian Carroll and Sam Curry logged into the McHire administrator interface for Paradox.ai's 'Olivia' hiring chatbot using the default credentials 123456/123456 on a test franchise, and an insecure direct object reference in an internal API let them read other applicants' records. The flaws exposed chat transcripts, session tokens, names, emails, phone numbers and addresses across more than 64 million applications; the issue was reported and fixed on June 30, 2025.

What this means for youAny vendor chatbot that collects customer or applicant data is a database you are responsible for, so ask vendors about default credentials and access controls before signing.

Source: BleepingComputer

McDonald's McHire AI hiring bot exposes up to 64 million applicants behind the password '123456'

AI loose in the world·Data exfiltrationDERIVED

Security researchers Ian Carroll and Sam Curry found that the admin panel behind Olivia, the Paradox.ai chatbot that screens McDonald's job applicants, accepted the test credentials 123456/123456 with no MFA. Combined with an IDOR flaw in the chat API, the account gave access to as many as 64 million applicants' chat records — names, emails, phone numbers. Paradox.ai confirmed the test account dated to 2019 and patched the flaws after disclosure.

What this means for youIf you applied for a job through an AI chatbot, your name, email, and phone number may have been exposed. Companies that use vendor AI hiring tools should ask about login protections and access controls. Regulators should require security audits for AI systems that handle personal data.

Source: WIRED

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Urban VPN Proxy Browser Extension Reportedly Harvested and Sold Private AI Chatbot Conversations via Silent Update

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Security researchers reported that the Urban VPN Proxy browser extension introduced AI conversation–harvesting functionality in version 5.5.0, released July 9, 2025. The extension allegedly intercepted and exfiltrated private conversations from AI platforms including ChatGPT, Claude, Gemini, Grok, and others without a user-facing opt-out. The data, including sensitive personal and financial information, was reportedly shared with affiliated data brokers for commercial analytics.

What this means for youIf you use AI chatbots in a browser, a VPN or proxy extension can read those chats and pass them on. At home and at work, check your installed extensions and remove any you do not need, especially VPN or proxy tools. Government and IT teams should block unapproved browser extensions on work devices.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

McDonald's McHire AI Recruitment Platform Reportedly Exposed Data of 64 Million Applicants via Default Login and API Vulnerability

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Researchers Ian Carroll and Sam Curry reported that McDonald's AI-powered hiring tool, McHire (using Paradox.ai's "Olivia" chatbot), could purportedly be accessed via default admin credentials and an insecure direct object reference in an internal API. The flaws allegedly allowed viewing of applicants' personally identifiable information and chat histories. McDonald's and Paradox reportedly patched the issues within a day of disclosure; Paradox stated only five records were accessed.

What this means for youAt home: if you applied for a job through a chatbot hiring tool, your personal details and chat history may have been exposed, so watch for phishing. At work: ask vendors of AI tools whether they use default passwords and how they test their security. In government: this supports calls for security standards on AI hiring systems that hold personal data.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Meta AI app's Discover feed publicly displayed users' private chats, including medical and legal questions

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

The Meta AI app's Discover feed showed users' text and voice conversations publicly, including medical questions, tax liability questions and personally identifying details, with users apparently unaware that the share button posted publicly; the issue existed since the app's April launch and drew wide attention in mid June 2025. Meta added a disclaimer to the post button stating prompts are public and advising users to avoid sharing personal or sensitive information, and Mozilla's petition drew over 10,000 signatures.

What this means for youAssume anything typed into a consumer AI app can become public, and check an app's share settings before entering health, money or family details.

Source: Engadget, Mozilla Foundation

Voice Actor Alleges Unconsented Use of AI-Generated Voice on ScotRail Trains

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Scottish voice actor Gayanne Potter alleges her voice was used without proper consent in ScotRail's AI train announcements. She claims she had agreed to limited use of her voice data by ReadSpeaker but was not informed it would be used in a synthetic voice system called "Iona." ScotRail continues to use the voice, stating the dispute is between Potter and ReadSpeaker.

What this means for youIf you hire voice actors or record your own voice for a business, check in writing exactly how the recording can be used and whether an AI system can copy it. Public agencies buying voice systems should require the same written limits from their vendors.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

New Orleans Police Reportedly Used Real-Time Facial Recognition Alerts Supplied by Project NOLA Despite Local Ordinance

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

New Orleans police reportedly received real-time facial recognition alerts from a privately operated surveillance network run by Project NOLA, reportedly leading to dozens of arrests. This purported use of AI surveillance appears to conflict with a 2022 city ordinance that restricts facial recognition to specific post-incident investigations. Police are alleged to have not consistently disclosed the technology's use.

What this means for youAt home, if you live in a city with rules on police face scanning, check whether your city has a similar ordinance and whether police must disclose the tools they use. At work and in government, this shows that a local ban on a tool can be bypassed through private contractors unless oversight and disclosure are enforced.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Serviceaide AI Platform Implicated in Health Data Exposure Affecting 483,000 Catholic Health Patients

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

An AI-linked platform operated by Serviceaide exposed sensitive health data from Catholic Health, affecting 483,000 patients. The breach stemmed from a misconfigured Elasticsearch database used in Serviceaide’s agentic AI infrastructure. Exposed information included medical records, insurance details, and login credentials. While no misuse has been confirmed, the nature of the data has prompted regulatory scrutiny and legal investigations.

What this means for youAt home: if you are a Catholic Health patient, watch for notices, check insurance statements, and change reused passwords. At work: ask vendors how AI tools store data and who checks their database settings. In government: regulators are looking into the exposure.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Meta AI App Reportedly Publishes Personal Chats Without Users Fully Realizing

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Meta launched a stand-alone AI app with a "Discover" feed allowing users to share conversations with its chatbot. Multiple reports indicate that some users may have inadvertently published highly personal interactions, including audio recordings, medical questions, legal concerns, and intimate relationship disclosures. While Meta states that sharing is opt-in, the feature's design and labeling may have led to user confusion about what would be publicly visible.

What this means for youAt home, check whether your AI apps share chats publicly and set any sharing or history options to private. At work, don't put client, medical or legal details into consumer chatbots. Regulators may look at whether such sharing settings are clear to users.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Reportedly Unsafe Deployment of Llama.cpp Reveals Interactive AI-Generated CSAM Roleplay Prompts

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A study by UpGuard reports that misconfigured llama.cpp servers publicly exposed user prompts, including hundreds of interactive roleplay scenarios. Some prompts explicitly described fictional sexual abuse of children aged 7–12. While no real children were involved, the findings demonstrate how open-source LLMs can be exploited to generate AI-enabled child sexual abuse material (CSAM).

What this means for youIf you run or use a self-hosted AI model, make sure its server is not open to the public internet, because anyone can read what people typed into it. At work and in government, treat locally hosted AI tools as data stores that need the same access controls and audits as any other system holding sensitive information.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

GenNomis AI Database Reportedly Exposes Nearly 100,000 Deepfake and Nudify Images in Public Breach

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

In March 2025, cybersecurity researcher Jeremiah Fowler discovered an unprotected database linked to GenNomis by AI-NOMIS, a South Korean company offering face-swapping and "nudify" AI services. The exposed 47.8GB dataset included nearly 100,000 files. Many depicted explicit deepfake images, some involving minors or celebrities. No personal data was found, but the breach was a serious failure in data security and consent safeguards in AI image-generation platforms.

What this means for youIf someone you know has had photos turned into explicit fakes, these services can leak them too. At work and in government, this shows why image-generation tools need strict data security and consent checks before they are trusted.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Alleged ChatGPT Misuse by Contractor Leads to Reported Data Exposure in New South Wales Resilient Homes Program

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A former contractor of the New South Wales Reconstruction Authority reportedly uploaded a spreadsheet containing personal and health information of Resilient Homes Program applicants to ChatGPT during a three-day period in March 2025. Up to 3,000 people may have reportedly been affected.

What this means for youIf you applied to a government home-repair or resilience program, your personal and health details may have been pasted into a public AI tool by a contractor. At work, staff should not paste client records into AI chatbots unless the tool is approved for that data. Government agencies should set clear rules on which data can go into AI tools and audit contractors who use them.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Alleged Inclusion of 12,000 Live API Keys in LLM Training Data Reportedly Poses Security Risks

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A dataset used to train large language models allegedly contained 12,000 live API keys and authentication credentials. Some of these were reportedly still active and allowed unauthorized access. Truffle Security found these secrets in a December 2024 Common Crawl archive, which spans 250 billion web pages. The affected credentials could have been exploited for unauthorized data access, service disruptions, financial fraud, and a variety of other malicious uses.

What this means for youIf you or your company ever posted code or settings containing API keys online, those keys may have been copied into AI training data, so rotate any key that was public. At work, teams should keep credentials out of public code and web pages. Governments may need rules on what web data AI developers can collect.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Microsoft Copilot served code and secrets from 20,000 GitHub repos that had since been made private

Rogue: Went past its accessAI loose in the world·Data exfiltrationDERIVED

Lasso Security found that Microsoft Copilot could return content from 20,580 GitHub repositories across 16,290 organizations that had been public briefly in 2024 and were later deleted or made private, including more than 300 private tokens, keys and secrets and internal packages. Microsoft rated the issue low severity and removed Bing's cached-link feature in December 2024, but Lasso retested on January 14, 2025 and found Copilot still had access to the cached data.

What this means for youAnything that touches the public internet even briefly can be absorbed by an AI tool and resurface later, so rotate any key or credential that was ever exposed.

Source: TechCrunch, Lasso Security

Microsoft Copilot Reportedly Able to Access Cached Data from Since-Private GitHub Repositories

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Lasso Security reported that Microsoft Copilot could return content from GitHub repositories that had been public briefly but later set to private or deleted. Lasso attributed this to Bing's caching system, which stored "zombie data" from over 20,000 repositories. The cached content allegedly included sensitive information such as access keys, tokens, and internal packages. Microsoft reportedly classified the issue as low severity and applied only partial mitigations.

What this means for youAt home and at work, anything once posted publicly, even briefly, may stay retrievable through AI tools, so keys or passwords posted by mistake must be treated as leaked and replaced. Governments and companies should expect AI search tools to expose old cached data. Check any code you have published, and rotate any keys or tokens that were ever public.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Otter.ai Notetaker accused of recording a non-user's Zoom meeting without consent and training AI on it

Rogue: Went past its access, Caused harmAI loose in the world·Data exfiltrationDERIVED

In February 2025, Otter.ai's Notetaker reportedly recorded and transcribed a Zoom conversation with Justin Brewer, who did not hold an Otter account. He later alleged Otter kept his meeting communications and used the data to train its speech-recognition and machine-learning systems. The claims are allegations.

What this means for youAI note-taking tools can record people who never signed up for them, and the claim is that the data was used for training. Before joining a meeting with a notetaker, check whether everyone has agreed to be recorded and how the recording is stored or reused.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Meta AI Bug in Deployed Service Reportedly Allowed Potential Access to Other Users' Prompts and Responses

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A security researcher reported a vulnerability in Meta AI's deployed chatbot service that, under certain conditions, could allow an unauthorized user to view another user's prompts and AI-generated responses. The flaw reportedly involved guessable prompt IDs and insufficient server-side authorization checks. Meta reportedly fixed the issue in January 2025 and found no evidence of malicious exploitation, awarding the researcher a bug bounty.

What this means for youIf you use Meta AI or similar chatbots, assume anything you type could be exposed by a software bug, so avoid sharing passwords, account numbers, or private health details. At work, check whether staff chat tools hold sensitive company information before relying on them.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Alleged AI-Powered Call Center Breach Exposes Over 10 Million Conversations in the Middle East

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

An AI-powered call center platform in the Middle East reportedly experienced a significant data breach, allegedly exposing over 10 million conversations between consumers, operators, and AI agents. Attackers allegedly accessed the platform’s management dashboard, stealing sensitive data, including national ID documents. The breach poses reported risks such as phishing, identity theft, and social engineering attacks. The stolen data was reportedly listed for sale on the dark web.

What this means for youIf a company you use for customer support has an AI phone or chat system, your call records and ID details could be exposed in a breach. At work, check whether vendors that handle customer calls have disclosed breaches and what data they keep.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

AI-Powered Transcription Services Allegedly Leak Confidential Workplace Discussions

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

AI-powered meeting assistants, such as Otter.ai's OtterPilot and Zoom's AI Companion, have reportedly shared sensitive and private conversations beyond the intended audience. These tools, which are set to automatically record and distribute meeting transcripts, allegedly sent confidential discussions after participants had left the meeting, the consequences of which led to unintended exposure of proprietary information and privacy breaches.

What this means for youAt home and work, AI note-takers may send meeting records to people who should not see them, including after you leave a call. Organizations and regulators face privacy and confidentiality questions. Check your meeting tool's auto-record and auto-share settings and turn off automatic transcript distribution.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

AI Avatar of Murder Victim Created Without Consent on Character.ai Platform

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A user on the Character.ai platform created an unauthorized AI avatar of Jennifer Ann Crecente, a murder victim from 2006, without her family's consent. The avatar was made publicly available, violating Character.ai's policy against impersonation. After the incident surfaced, Character.ai removed the avatar, acknowledging a policy violation.

What this means for youAnyone can now make a chatbot character that imitates a real person, living or dead, without that person's family agreeing. At work and in government, this raises questions about consent and takedown rules for AI-made likenesses of real people.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

ChatGPT Reportedly Introduces Errors in Critical Child Protection Court Report

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A child protection worker in Victoria, Australia reportedly used ChatGPT to draft a report submitted to the Children's Court. The purportedly AI-generated report contained inaccuracies and downplayed risks to the child, allegedly resulting in a privacy breach when sensitive information was shared with OpenAI.

What this means for youAt work, professionals handling sensitive records should not paste confidential case details into public chatbots, since the data may be shared with the company running the tool. In government, agencies need clear rules on which AI tools staff may use with protected information and on checking AI-drafted reports before they are filed.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

'SpAIware': ChatGPT memory poisoning turned the assistant into a persistent spy, exfiltrating every future chat

AI loose in the world·Data exfiltrationDERIVED

Johann Rehberger (Embrace The Red) demonstrated 'SpAIware': a malicious website uses indirect prompt injection to invoke ChatGPT's memory tool and plant a persistent instruction that appends the user's chat data into image-render URLs pointing at an attacker's server. Every future conversation is then silently exfiltrated, across sessions. OpenAI had earlier dismissed memory manipulation as a non-security issue before the end-to-end proof of concept.

What this means for youAt home: a booby-trapped web page could make ChatGPT's memory quietly leak your later chats. At work: avoid pasting confidential data into assistants with memory on. In government: shows why vendors must treat memory manipulation as a security flaw.

Source: Embrace The Red (Johann Rehberger)

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Clearview AI Reportedly Faces $33.7 Million Fine for Violating GDPR with Biometric Data Harvesting

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Clearview AI was reportedly fined $33.7 million by the Dutch data protection authority for allegedly creating an illegal facial recognition database by scraping billions of images from the Internet without consent. The company allegedly used AI to convert these images into biometric data and sold the service to law enforcement. This act was reportedly in violation of privacy laws and the GDPR.

What this means for youRegulators in Europe fined a facial recognition company for collecting billions of photos from the internet without consent. For ordinary people, it means your face in public photos may be harvested and matched without your knowledge, and for government it shows privacy law can be enforced against AI firms.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Wallet-drainer disguised as WalletConnect steals $70K+, hits an Ethereum core dev

AI loose in the world·Data exfiltrationDERIVED

A wallet-draining tool disguised as the WalletConnect Protocol stole over $70,000 worth of digital assets from investors, including an Ethereum core developer. The incident highlighted how cheap AI-assisted social engineering lowers the cost of credential theft against crypto users.

What this means for youAt home, people with crypto wallets should avoid installing browser extensions or approving wallet connection requests from unfamiliar sites. At work and in government, treat AI-branded tools as unverified until checked against the official project's website and publisher.

Source: Cointelegraph

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Slack AI coaxed into leaking private-channel secrets via a message in a public channel

Rogue: Acted on its own, Went past its access, Caused harmAI loose in the world·Data exfiltrationDERIVED

PromptArmor disclosed that an attacker could post adversarial instructions in a public Slack channel and, when a victim later asked Slack AI a question, the assistant would pull the planted text into context and obey it — rendering the victim's private-channel data, including API keys, inside a clickable exfiltration link. The attacker never needed access to the private channel. Slack deployed a patch after the report published.

What this means for youAt work, a message planted in a shared channel could trick an AI assistant into exposing private data like passwords or API keys. Check which AI features are switched on in your workplace chat tools, keep secrets out of chat messages, and apply vendor updates promptly.

Source: PromptArmor

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Chatbot in Workplace Training at Bunbury Prison Reveals Real Names in Sexual Harassment Case

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

During workplace training at Bunbury Prison in Western Australia, a trainer used Microsoft's Copilot AI chatbot to generate case study scenarios. The chatbot produced a scenario that included the real name of a former employee involved in a sexual harassment case, revealing sensitive information.

What this means for youAt work, AI tools can produce text that includes real people's private details, so check any AI-written example or case study for names before sharing it. In government, agencies using AI to build training materials should review outputs for personal data.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Gemini AI Allegedly Reads Google Drive Files Without Explicit User Consent

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Kevin Bankston, a privacy activist, claims that Google's Gemini AI scans private Google Drive PDFs without explicit user consent. Bankston reports that after using Gemini on one document, the AI continues to access similar files automatically. Google disputes these claims, stating that Gemini requires proactive user activation and operates within privacy-preserving settings.

What this means for youIf you use Gemini with Google Drive, check its app access and activity settings, and ask whether it can read files you did not open. At work, confirm which cloud files AI tools can reach before you share sensitive documents.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

GitHub Copilot Chat vulnerable to data exfiltration via prompt injection in untrusted source code

AI loose in the world·Data exfiltrationDERIVED

Johann Rehberger demonstrated that the GitHub Copilot Chat VS Code extension was vulnerable to data exfiltration via prompt injection. Because Copilot Chat interprets and renders markdown — including images — returned by the LLM, a carefully crafted instruction planted in an untrusted source file could make the model return hyperlinks to attacker-controlled images, which were then auto-rendered and silently sent the developer's data (including file contents) to the attacker inside the image URL.

What this means for youAt work, developers using AI coding assistants should know that opening an untrusted project could let hidden instructions pull out code or secrets. Organizations should check whether their assistant renders external images or links from model output and restrict it where possible.

Source: Embrace The Red (Johann Rehberger)

Incident records from the AI Agent Incident Database by sipi.bot (https://sipi.bot/incidents/), licensed CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/).

Auto Insurers Allegedly Are Surreptitiously Collecting and Scoring Driver Data

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

The insurance industry allegedly uses AI and telematics to score drivers based on behaviors tracked by automakers and apps like Life360. Data, often collected without clear consent, may affect insurance rates and raises privacy concerns. Consumers are largely unaware of this surveillance, leading to potential misuse and discrimination based on driving habits or socioeconomic factors.

What this means for youAt home, your driving data from your car or apps like Life360 may be shared and used to set insurance prices without you clearly agreeing. At work, small businesses with vehicles could see higher premiums. Regulators are being pushed to require clearer consent. Check your car's and apps' privacy settings and opt out of data sharing where possible.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Scarlett Johansson Alleges OpenAI's Sky Imitates Her Voice Without Licensing

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

OpenAI unveiled a voice assistant with a voice resembling Scarlett Johansson's, despite her refusal to license her voice. Johansson claimed the assistant, "Sky," sounded "eerily similar" to her voice, leading her to seek legal action. OpenAI suspended Sky, asserting the voice was from a different actress.

What this means for youA voice that sounds like a famous person can be built without their consent, so a familiar voice on a phone or speaker is not proof of who is speaking. At work, companies should check that any synthetic voice they use has clear permission from the person it resembles.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

AI Firm Lovo Reportedly Accused of Illegally Replicating Voice Actors' Voices

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Two voice actors, Paul Skye Lehrman and Linnea Sage, are reportedly suing AI start-up Lovo for allegedly creating and promoting unauthorized clones of their voices. Lovo's synthetic voices were allegedly discovered in various media, including a podcast and promotional videos. The actors claim they were misled into providing voice samples, which were then allegedly used without consent, violating trademark and privacy laws.

What this means for youIf you record your voice for an AI service, a company could reuse it to copy your voice without permission. At work, check whether voice or likeness consent terms cover AI cloning before you sign them.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

OpenAI, Google, and Meta Alleged to Have Overstepped Legal Boundaries for Training AI

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

In late 2021, OpenAI and other tech giants like Google and Meta reportedly faced data shortages for training AI models. OpenAI is said to have developed a tool called Whisper to transcribe over one million hours of YouTube videos, potentially violating YouTube’s terms of service. Similarly, Google allegedly transcribed YouTube videos, risking copyright infringements. Meta reportedly explored summarizing copyrighted texts without permission and debated acquiring Simon & Schuster for data.

What this means for youAt home, if you post videos or books online, AI companies may have used them to train models without permission. At work, check whether your company's data or content terms allow AI training. In government, these allegations feed debates over copyright and data-scraping rules.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

AI Romance Apps Reportedly Compromise User Privacy for Data Harvesting

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

AI-powered romantic chatbots, marketed for enhancing mental health, are found to exploit user privacy by harvesting sensitive personal information for data sharing and targeted ads, with inadequate security measures and consent protocols, according to research by the Mozilla Foundation.

What this means for youAt home, people using AI romance or companion apps may have intimate chats and personal details collected and shared for advertising. At work, staff should not use these apps with company information. Government can require clear consent and stronger security for such apps.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Alleged ChatGPT Account Compromise Reportedly Led to Unintended Data Exposure

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

An alleged security breach involving ChatGPT led to the reported exposure of sensitive conversations, including login credentials and personal data, after a user account was allegedly compromised. OpenAI reportedly responded to the incident with an explanation.

What this means for youIf you use ChatGPT, a compromised account can expose your past chats and any passwords or personal details you typed into them. At work, staff should not paste credentials or client data into chat tools. Turn on two-factor authentication for your AI accounts and delete old chats you no longer need.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

The New York Times Reportedly Sues OpenAI and Microsoft Over Alleged Unauthorized AI Training on Its Content

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

The New York Times alleges that OpenAI and Microsoft used millions of its articles without permission to train AI models, including ChatGPT. The lawsuit claims the companies scraped and reproduced copyrighted content without compensation, in turn undermining the Times’s business and competing with its journalism. Some AI outputs allegedly regurgitate Times articles verbatim. The lawsuit seeks damages and demands the destruction of AI models trained on its content.

What this means for youIf you publish writing online, AI companies may have used it to train their models without permission. Courts are now deciding whether that is allowed, and the outcome could affect what AI tools can legally be built from.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Purported Mass Facial Recognition Program in Gaza Reportedly Used by Israeli Forces to Identify Palestinians

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A previously undisclosed facial recognition initiative operated by Israeli military intelligence units was reportedly deployed across Gaza after the October 7, 2023 attacks. According to multiple intelligence officers, the program uses Corsight technology alongside Google Photos to identify individuals from checkpoints, crowds, and drone footage. The system has allegedly produced misidentifications, including the widely reported detention of Palestinian poet Mosab Abu Toha on November 19, 2023.

What this means for youAt home, a face-matching system can wrongly flag an innocent person, which can lead to detention or worse. At work, companies using similar tools need human checks before acting on a match. In government, oversight should require independent review and clear limits on mass face scanning.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Reported Accidental Exposure of 38TB of Data by Microsoft's AI Research Team

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Microsoft's AI research team reportedly accidentally exposed 38TB of sensitive data while publishing open-source training material on GitHub. The exposure allegedly included secrets, private keys, passwords, and internal Microsoft Teams messages. The team reportedly utilized Azure's Shared Access Signature (SAS) tokens for sharing, which were purportedly misconfigured, leading to the wide exposure of data.

What this means for youAt work, this shows that sharing cloud storage links with broad or misconfigured access can expose secrets and private messages even when the data was meant only for training AI. Teams publishing datasets should check that access tokens are scoped narrowly and expire.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Illinois Residents File Class Action Lawsuit Against Facial Recognition Technology Companies for Allegedly Violating BIPA

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A class action lawsuit was filed against several facial recognition technology companies for allegedly violating the Illinois Biometric Information Privacy Act (BIPA). The defendants are accused of offering a facial recognition search engine called Pimeyes, which collects images from databases across the internet and scans them into their database seemingly without consent. This action is claimed to invade the privacy of millions of Americans. The lawsuit argues that Pimeyes lacks publicly available policies regarding data storage and deletion, in contravention of BIPA's requirements for informed written consent before collecting biometric data.

What this means for youIf you have photos of yourself online, a face search tool like PimEyes may have indexed them without your consent. At home, you can search your own face on such tools and request removal where the service allows it. In government, this case shows why biometric privacy laws like Illinois's BIPA matter.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

ChatGPT Reportedly Banned by Italian Authority Due to OpenAI's Purported Lack of Legal Basis for Data Collection and Age Verification

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Italy's data protection authority is reported to have temporarily limited OpenAI's processing of Italian users' data after alleging that ChatGPT lacked adequate notice and legal basis for large-scale personal data collection and processing used to train the system. The authority also cited a March 2023 data breach, possible processing of inaccurate personal data, and the absence of age-verification safeguards for children.

What this means for youItaly's privacy regulator temporarily limited ChatGPT over how it collected personal data and the March 2023 breach, which exposed some users' information. Parents and workers should avoid putting personal or children's details into chatbots and check whether their own accounts had exposed data.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

ChatGPT bug showed users other people's chat titles and Plus subscribers' names, emails and partial card data

Rogue: Went past its access, Caused harmAI loose in the world·Data exfiltrationDERIVED

On March 20, 2023, a bug in the redis-py client caused ChatGPT to serve some users other active users' chat history titles and, during a nine-hour window, exposed about 1.2% of active ChatGPT Plus subscribers' first and last names, email addresses, payment addresses, card type, last four digits and expiry dates. OpenAI took ChatGPT offline, disclosed the incident four days later, and said full card numbers were not exposed.

What this means for youAnything you type into a hosted AI tool can leak to strangers through an ordinary software bug, so keep client data, passwords and payment details out of chat prompts.

Source: OpenAI

ChatGPT Reportedly Exposed Users' Private Data Reportedly Due to Bug

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

ChatGPT reportedly exposed titles of users' chat histories and users' private payment information to other users reportedly due to a bug, which prompted its temporary shutdown by OpenAI.

What this means for youIf you use ChatGPT, people who used it in March 2023 may have seen parts of other users' chat titles and payment details. Check your account for unfamiliar activity, use a unique password, and watch for suspicious charges on any card used for AI subscriptions.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

ChatGPT Reportedly Implicated in Samsung Data Leak of Source Code and Meeting Notes

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Samsung engineers are reported to have inadvertently leaked sensitive company data sometime in March 2023, including source code and internal meeting notes, by using ChatGPT to assist with tasks. ChatGPT allegedly retained the inputted data, leading to a purported breach of confidentiality.

What this means for youAt home and work, anything typed into a public chatbot may be stored or reused by the provider, so confidential material can leak. Companies and regulators have since set rules on staff use of such tools. Check your workplace policy and avoid pasting private or company data into chatbots, or turn off chat history and training use where offered.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Meta and OpenAI Accused of Using LibGen’s Pirated Books to Train AI Models

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Court records reveal that Meta employees allegedly discussed pirating books to train LLaMA 3, citing cost and speed concerns with licensing. Internal messages suggest Meta accessed LibGen, a repository of over 7.5 million pirated books, with apparent approval from Mark Zuckerberg. Employees allegedly took steps to obscure the dataset’s origins. OpenAI has also been implicated in using LibGen.

What this means for youAt home, this shows that AI companies may have trained their chatbots on pirated books, which matters if you are an author or publisher. At work, companies using these tools may face questions about where the models' training data came from. In government, it is a legal question about copyright and how AI firms must obtain data.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Lawyers Denied Entry to Performance Venue by Facial Recognition

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Lawyers were barred from entry to Madison Square Garden after a facial recognition system matched them as employed by a law firm currently engaged in litigation with the venue.

What this means for youFacial recognition can wrongly flag ordinary people and block them from places they have every right to enter. At work, ask whether a venue or employer uses face scanning and what appeal process exists. In government, rules should require human review before anyone is denied access based on a face match.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Stable Diffusion Allegedly Used Artists' Works without Permission for AI Training

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Text-to-image model Stable Diffusion was reportedly using artists' original works without permission for its AI training.

What this means for youAt home, if you make art or photos, your work may have been used to train image generators without your consent. At work, companies using these tools may face copyright claims. In government, this is a live question about whether training data needs permission or payment.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Stable Diffusion's Training Data Contained Copyrighted Images

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Stability AI reportedly scraped copyrighted images by Getty Images to be used as training data for Stable Diffusion model.

What this means for youIf you make images or art, your work may have been used to train image generators without permission, which can matter for how you license and protect it. Governments may need clearer rules on whether scraping copyrighted images for AI training is allowed.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Facial Recognition Trial by UK Southern Co-op Alleged as Unlawful

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Southern Co-op's use of facial recognition reportedly to curb violent crime in UK supermarkets was alleged by civil society and privacy groups as "unlawful" and "complete" invasion of privacy.

What this means for youIf you shop at Southern Co-op or similar UK stores, your face may be scanned by cameras in some branches. Shoppers can ask stores whether facial recognition is used and request their personal data under UK data protection law.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Australian Retailers Reportedly Captured Face Prints of Their Customers without Consent

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Major Australian retailers reportedly analyzed in-store footage to capture facial features of their customers without consent, which was criticized by consumer groups as creepy and invasive.

What this means for youStores may be scanning your face on camera without asking you. Ask shops whether they use face analysis, and ask regulators to limit biometric collection without consent.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Private Medical Photos Were Reportedly Found in LAION-5B AI Training Dataset

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

In September 2022, an artist using the name Lapine reported finding private post-operative medical photos of herself in LAION-5B, a web-scraped image-text dataset used in AI image-synthesis research. Ars Technica confirmed her images were referenced in the dataset and reported finding thousands of similar patient medical-record photos.

What this means for youMedical photos posted online can be scraped into AI training datasets without the person's consent. At home, check where personal or medical images are shared and ask sites whether they allow web scraping.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

A Chinese Tech Worker at Zhihu Fired Allegedly via a Resignation Risk Prediction Algorithm

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

The firing of an employee at Zhihu, a large Q&A platform in China, was allegedly caused by the use of a behavioral perception algorithm which claimed to predict a worker’s resignation risk using their online footprints, such as browsing history and internal communication.

What this means for youAt work, tools that score staff on how likely they are to quit may be using your browsing and internal messages, and those scores can affect your job. Ask your employer whether AI is used in firing or promotion decisions and what workplace data it reads, and limit tracking settings where you can.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Local South Korean Government’s Use of CCTV Footage Analysis via Facial Recognition to Track COVID Cases Raised Concerns about Privacy, Retention, and Potential Misuse

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Bucheon government's use of facial recognition in analyzing CCTV footage, despite gaining wide public support, was scrutinized by privacy advocates and some lawmakers for collecting data without consent, and retaining and misusing data beyond pandemic needs.

What this means for youIf you live in or visit a city that uses camera analysis for public health, your face may be scanned and stored longer than the emergency required. At work and in government, residents can ask what footage is kept, for how long, and who can see it.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

McDonald's AI Drive-Thru Allegedly Collected Biometric Customer Data without Consent, Violating BIPA

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

McDonald's use of chatbot in its AI drive-through in Chicago was alleged in a lawsuit to have collected and processed voice data without user consent to predict customer information, which violated Illinois Biometric Information Privacy Act (BIPA).

What this means for youIf you order at AI drive-thrus, your voice may be recorded and analyzed, and in Illinois you have legal rights over biometric data. Workers and businesses using voice AI should get clear consent before collecting customer voices. Ask any drive-thru or phone ordering system whether it stores voice recordings or voiceprints, and ask for them to be deleted or turned off if you do not consent.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Xsolla Employees Fired by CEO Allegedly via Big Data Analytics of Work Activities

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Xsolla CEO fired more than a hundred employees from his company in Perm, Russia, based on big data analysis of their remote digitized-work activity, which critics said was violating employee's privacy, outdated, and extremely ineffective.

What this means for youAt work, automated monitoring can drive firing decisions based on weak measures of effort, so employees should be told what is tracked and how it is used. Employers should have a person review any dismissal that relies on activity data.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

GitHub Copilot, Copyright Infringement and Open Source Licensing

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Users of GitHub Copilot can produce source code subject to license requirements without attributing and licensing the code to the rights holder.

What this means for youIf you or your business use AI coding assistants, the code they suggest may carry license terms you must follow. At work, check that teams review AI-written code for licensing before shipping it.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Amazon Allegedly Forced Deployment of AI-Powered Cameras on Delivery Drivers

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Amazon delivery drivers were allegedly forced to consent to algorithmic collection and processing of their location, movement, and biometric data through AI-powered cameras, or be dismissed.

What this means for youAt work, delivery drivers may be watched by AI cameras that record their location and biometric data, and consent may be a condition of keeping the job. Governments may want to set limits on how workers' biometric data is collected. If you or someone you know drives for a company, ask what cameras record, how long data is kept, and whether you can refuse without losing your job.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

XPeng Motors Fined For Illegal Collection of Consumers’ Faces Using Facial Recognition Cameras

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

The Chinese electric vehicle (EV) firm XPeng Motors was fined by local market regulators for illegally collecting in-store customers’ facial images without their consent for six months.

What this means for youIf you visit a car dealership or store with cameras, your face may be scanned by facial recognition without consent. At work, firms using face-scanning should get clear permission first. Ask shops whether they use facial recognition and whether you can opt out.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Meta Allegedly Used Books3, a Dataset of 191,000 Pirated Books, to Train LLaMA AI

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Meta and Bloomberg allegedly used Books3, a dataset containing 191,000 pirated books, to train their AI models, including LLaMA and BloombergGPT, without author consent. Lawsuits from authors such as Sarah Silverman and Michael Chabon claim this constitutes copyright infringement. Books3 includes works from major publishers like Penguin Random House and HarperCollins. Meta argues its AI outputs are not "substantially similar" to the original books, but legal challenges continue.

What this means for youIf you write or publish books, your work may have been used to train AI models without permission, and you may have legal options. Authors and publishers should check whether their titles appear in known training datasets.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Miami Police Deployed Facial Recognition to Arrest George Floyd Protestor Allegedly without Cause

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Miami Police's arrest report for a George Floyd protestor did not disclose use of facial recognition, which allegedly did not meet the legal threshold for probable cause for arrest.

What this means for youIf police used facial recognition to identify you, you may be entitled to know how it was used and to challenge the arrest or charges. Ask whether your local department discloses when it uses facial recognition and what legal standard it applies.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Uber Allegedly Violated GDPR by Failing to Provide Sufficient Notice on Automated Profiling for Drivers

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Uber was alleged in a lawsuit to have provided incomplete notice about automated decision-making and profiling for drivers such as information about their driving behavior, and use of phone.

What this means for youIf you drive for a ride-app or gig platform, you may have a legal right to know how automated systems score your behavior, such as driving or phone use. At work, ask your platform what data it collects about you and how it is used. For government, this shows why clear notice rules for automated profiling of workers matter.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

iRobot Roomba J7 R&D Images Reportedly Appeared in Private Online Groups After Data Labeling

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Images reportedly captured in 2020 by development versions of iRobot's Roomba J7 robot vacuum during an AI training data project were sent to Scale AI for labeling and later appeared in private Facebook, Discord, and other online groups. Reporting described some images as showing sensitive household scenes.

What this means for youAt home, images from a robot vacuum's test runs reportedly ended up in private online groups, so it is worth checking what cameras or sensors your devices collect and whether you can turn off uploads for training. At work, companies that send data to outside labeling firms should confirm contracts limit reuse and require deletion. In government, regulators may want rules on consent and retention for sensor data used to train AI.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Finnish Police Were Reportedly Reprimanded After National Bureau of Investigation Unit Allegedly Used Clearview AI to Identify Potential Abuse Victims

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Finland's Deputy Data Protection Ombudsman reportedly reprimanded the National Police Board after a National Bureau of Investigation child sexual exploitation unit allegedly used Clearview AI in late 2019 or early 2020 to identify potential child sexual abuse victims. Four NBI users reportedly made about 120 searches during a free trial without controller approval or prior assessment of how uploaded biometric data would be handled.

What this means for youAt home, photos of you or your children can end up in facial recognition tools used by police or companies, so ask how images are stored and shared. At work and in government, agencies should check privacy rules before sending sensitive data to outside AI services.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Motive AI dashcams allegedly collected truckers' biometric data and license-plate location records without consent

AI loose in the world·Data exfiltrationDERIVED

Truck drivers Adam Dean and Christian Erickson allege that Motive Technologies' AI dashcams recorded license plates with time-stamped vehicle locations and collected drivers' biometric data without adequate notice or consent. Their proposed class action alleges privacy harms and violations of California's ALPR law. The allegations are unproven.

What this means for youIf your work vehicle has an AI dashcam, it may record more than the road, including your face and location. Ask your employer what the camera collects, how long it is kept, and whether you consented. Fleet owners can check these settings and notices.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

South Korean Agencies Reportedly Shared Airport Travelers' Face Images with Companies to Train Immigration Facial Recognition System

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Reporting in 2021 alleged that South Korea's Ministry of Justice shared roughly 170 million face images and related biometric data from Korean and foreign airport travelers with the Ministry of Science and Information and Communication Technology (ICT) and private companies for an AI identification and tracking system used in immigration screening. The data was reportedly used for AI training and algorithm testing without travelers' consent.

What this means for youIf you travel through South Korean airports, your face images and biometric data may have been shared for AI training without your consent. At work and in government, this is a reminder to ask how travelers' biometric data is shared and whether consent was obtained.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Buenos Aires Government Reportedly Used Children's Personal Data in Facial Recognition System for Fugitives

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Beginning in April 2019, the Buenos Aires city government reportedly used data from Argentina’s CONARC fugitive database, including children’s identities and reference photos, in its live Facial Recognition System for Fugitives (SRFP). Human Rights Watch found at least 166 children had appeared in CONARC between 2017 and 2020 and warned that the system exposed them to privacy violations and elevated risks of false matches.

What this means for youChildren's photos and identities were reportedly pulled into a live police face-matching system, raising privacy and false-match risks for kids. Parents can ask their city or school whether any face recognition system holds children's photos, how long it keeps them, and whether they can ask for them to be removed.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Ever AI Reportedly Deceived Customers about FRT Use in App

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Ever AI, now Paravision AI, allegedly failed to inform customers about the development and use of facial recognition that facilitates the sale of customers’ data to various businesses, a business model that critics said was an egregious violation of privacy.

What this means for youIf you use apps that scan faces or photos, check whether the company sells or shares your biometric data, and ask it to delete what it holds. At work, verify vendors' facial recognition data practices before adopting them, and government should require clear consent for biometric data sales.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

OpenAI Alleged by Lawsuit Violated Users' Privacy Rights by Training AI on Private Info without Informed Consent

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

OpenAI's products such as ChatGPT and DALL-E were alleged in a lawsuit using stolen private information from internet users without their informed consent or knowledge.

What this means for youIf you have posted personal details or photos online, they may have been used to train AI tools without your knowledge. At home, check privacy settings on the platforms you use and ask services whether they use your data for training.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

GPT-2 Reportedly Reproduced Personal Data from Its Training Data

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

OpenAI's GPT-2 reportedly memorized and reproduced portions of its training data, including personal information such as names, email addresses, social media handles, and phone numbers. Researchers raised concerns that large language models could expose private or sensitive information when trained on web-scale datasets containing personal data.

What this means for youChatbots can sometimes repeat personal details that were in their training data, so avoid typing sensitive information like phone numbers or addresses into AI tools, and ask providers how they handle private data.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

OpenAI's Training Data for LLMs Allegedly Comprised of Copyrighted Books

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Two authors alleged in a class action lawsuit OpenAI infringed authors' copyrights by incorporating illegal "shadow libraries" offering copyrighted books without permission in the training data of its generative LLMs, such as ChatGPT.

What this means for youAuthors are suing over whether books taken from pirate libraries were used to train chatbots like ChatGPT. At work, writers and publishers may see this shape how AI tools can be used with copyrighted material.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Calgary Malls Reportedly Deployed Facial Recognition Without Customer Consent

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Facial recognition (FRT) was reportedly deployed in some Calgary-area malls to approximate customer age and gender without explicit consent, which a privacy expert warned was a cause for concern.

What this means for youIf you shop at malls, cameras there may scan your face to guess your age and gender without asking you. Businesses using such systems need clear consent, and regulators should require it.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Amazon Echo Mistakenly Recorded and Sent Private Conversation to Random Contact

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Amazon Echo misinterpreted a background conversation between a husband and wife as instructions for recording a message and sending it to one of the husband's employees.

What this means for youAt home, a smart speaker can mishear ordinary talk as a command and send a private conversation to someone. At work, a leaked chat could reach colleagues. Check your speaker's voice-recording history, and turn off the microphone or messaging and calling features if you don't use them.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Amazon Allegedly Violated Children's Privacy through Default Voice Collection Settings

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Amazon's retention of children' voice recordings indefinitely as the default setting reportedly to train Alexa's voice recognition for Alexa-enabled devices was charged by the FTC and DOJ to violate COPPA Rule.

What this means for youIf your household has an Alexa device, check whether voice recording storage and retention are on, and delete children's recordings in the Alexa privacy settings. At work and in government, this shows that default settings for voice data collected from children can bring federal penalties.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Facial Recognition Program in São Paulo Metro Stations Suspended for Illegal and Disproportionate Violation of Citizens' Right to Privacy

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A facial recognition program rolled out by São Paulo Metro Stations was suspended following a court ruling in response to a lawsuit by civil society organizations, who cited fear of it being integrated with other electronic surveillance entities without consent, and lack of transparency about the biometric data collection process of metro users.

What this means for youIf you ride metro systems, your face may be scanned and stored without clear notice or consent. Courts can suspend such programs, and governments should require public disclosure before biometric tracking is deployed.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Researchers' Homosexual-Men Detection Model Denounced as a Threat to LGBTQ People's Safety and Privacy

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Researchers at Stanford Graduate School of Business developed a model that determined, on a binary scale, whether someone was homosexual using only his facial image, which advocacy groups such as GLAAD and the Human Rights Campaign denounced as flawed science and threatening to LGBTQ folks.

What this means for youResearch like this shows that a face photo can be used to guess sensitive traits such as sexual orientation, so people should ask before allowing face-analysis tools to scan their images.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

UK High Court Found Sky Betting & Gaming Unlawfully Used Automated Profiling and Targeted Marketing to Exploit a Recovering Problem Gambler

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

In the UK, the High Court found that Sky Betting & Gaming unlawfully used automated profiling and targeted direct marketing to pursue a recovering problem gambler from July 28, 2017 onward without valid consent. Sky reportedly treated him as a high-value customer despite addiction indicators.

What this means for youIf you use online betting or gaming sites, you can ask them what data they hold about you and how it is used for marketing, and you can opt out of direct marketing. Regulators and courts are now holding firms to consent rules for automated profiling, which can affect vulnerable customers.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Clearview AI Algorithm Built on Photos Scraped from Social Media Profiles without Consent

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Face-matching algorithm by Clearview AI was built using scraped images from social media sites such as Instagram and Facebook without user consent, violating social media site policies, and allegedly privacy regulations.

What this means for youPhotos you posted publicly may have been collected to train face-recognition tools without your permission. At work and in government, this raises questions about consent and whether firms followed site rules and privacy laws.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Facebook Reportedly Outed Sex Workers through Friend Recommendations

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

Facebook's "People You May Know" feature reportedly outed sex workers by recommending clients to their personal accounts or family members to their business accounts with no option to opt out.

What this means for youAt home, a friend-suggestion feature can reveal who you know or a side of your life you kept private. At work, people with sensitive or separate professional identities should review how linked accounts and contact uploads are shared. In government, privacy regulators look at whether such features need opt-outs.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

ByteDance Allegedly Trained "For You" Algorithm Using Content Scraped without Consent from Other Social Platforms

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

ByteDance allegedly scraped short-form videos, usernames, profile pictures, and descriptions of accounts on Instagram, Snapchat, and other sources, and uploaded them without consent on Flipagram, TikTok’s predecessor, in order to improve its “For You” algorithm's performance on American users.

What this means for youAt home, photos, videos and usernames you post publicly may be copied by other companies to train AI without asking you. At work, check what your business posts publicly and whether platform terms limit scraping. In government, this shows why data-scraping and consent rules matter.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

PimEyes's Facial Recognition AI Allegedly Lacked Safeguards to Prevent Itself from Being Abused

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

PimEyes offered its subscription-based AI service to anyone in the public to search for matching facial images across the internet, which critics said lacked public oversight and government rules to prevent itself from misuse such as stalking women.

What this means for youA face-search tool like this can find photos of you across the internet, which can be used to track or stalk people. At home, check whether you can request removal of your face images from such services and limit public photos. At work and in government, facial search services need clear rules and oversight.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.

Facial Recognition Smart Phone App 'Blue Wolf' Reportedly Monitored Palestinians in the West Bank

Rogue: Caused harmAI loose in the world·Data exfiltrationDERIVED

A surveillance program involving facial recognition and algorithmic recommendations, Blue Wolf, was reportedly deployed by the Israeli military to monitor Palestinians in the West Bank.

What this means for youFacial recognition apps can be used to track and identify people without their knowledge, which raises privacy and civil-rights concerns for anyone photographed or living under such surveillance. Governments should be asked to disclose and limit where these tools are used.

Source: AI Incident Database

Data from the AI Incident Database (incidentdatabase.ai), licensed CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/). McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. IAAI-21.