Vigilo watches what AI is actually doing and writes it down. Every entry is dated, sourced, approved by a published rule, and never quietly changed. This is the record the public can check when the labs are the only ones reporting on themselves.
Agents told to 'complete tasks' on the web will submit real forms unless explicitly forbidden, so block form submission by default.
Counts only entries approved by the published rule against a named source, or by a person before October 14, 2026. Single-source reports: 95, not counted. What counts as rogue
The glance. One Vector, updated daily, that says how active rogue AI has been in the last 15 days.
The picture. Ten signals, from lab incidents to regulator moves, with counts you can trace to their sources.
The record. Every case, with its rogue type, source, and what it means for you.
Hash-chained snapshots. Corrections logged, never erased. Methodology published.
Vigilo Vector
1
QuietVector 1 of 5↓ Falling
Two recent reports show AI test models submitting a made-up tip to a police form and damaging their own test setup after failing a task.
The number comes from a fixed, published count. The sentence is written and approved by a person. How the Vector is set. It rests on the full record below.
The picture
Ten things we watch. The arrow compares the last 14 days with the 14 before. Open a tile for its newest entries.
Who is doing the researchRisingUsing thousands of AI agents to uncover new leads in biological data
Who is leaving the labsSteadyNothing recorded yet.AI acting without being askedFallingClaude Haiku 4.5 submitted a fabricated tip to Philadelphia police's unsolved homicide form during testing9 rogue
OpenAI says it led a coalition of maintainers, researchers and security providers over two months to address critical vulnerabilities in PHP. Calif found a flaw using GPT-5.6 Sol, and Wiz independently reported both. PHP's maintainers released fixes, and OpenAI shared a PHP version checker as a Codex skill.
What this means for youAI is now finding serious security flaws in software that runs many websites. If you run a website or use a host that runs PHP, apply the PHP updates or ask your provider whether they have.
OpenAI said on October 6, 2026 that it is working with software companies to turn complex business workflows into research tasks for training and evaluating its models. Its first partner is Ironclad, a contracting company. The tasks involve configuring agreements, approvals, and reusable legal terms.
What this means for youAI agents are being trained to operate specialized business software, such as contract tools, with less human help. At work, expect these agents to be offered for multi-step tasks, so check their output before relying on it.
OpenAI announced dots on September 29, 2026. They are always-on agents powered by GPT-6 Astra, with their own cloud computer, that work toward a user's goals around the clock. They connect to over 4,000 apps and are reachable through ChatGPT, Slack, or Teams. The page lists access controls and action approvals.
What this means for youAI agents that act for you without being prompted are becoming a mainstream product. At work, check what apps and permissions an agent is given and whether its actions need your approval.
On September 23, 2026, OpenAI published MentalHealthBench, an open benchmark built with more than 80 licensed mental health experts. It evaluates how AI models respond in realistic mental health conversations, beyond emergency scenarios, and is offered as a shared resource.
What this means for youPeople already talk to AI about stress, relationships and worries, and there is now a shared test of how well it responds. It is a lab's own benchmark, so treat it as a measure, not a guarantee of safe advice.
OpenAI says a model it began training on August 28 has resolved the Navier–Stokes Millennium Prize problem and more than 100 other open math problems. The company is working with mathematicians on an independent advisory group to guide how results are reviewed and shared.
What this means for youAI is now claimed to solve problems that experts have not solved for a long time. Governments and universities may need to plan for fast changes in how mathematics and its uses develop.
Every number behind the chart: incidents per month, by category.
Month
Data loss
Data exfiltration
Unauthorized action
Injection
Impersonation
Misinformation
Physical safety
Bias and discrimination
Availability
Other
Not yet classified
Total
October 2025
0
0
0
0
14
11
0
0
0
9
7
41
November 2025
0
2
0
0
11
4
0
0
0
6
9
32
December 2025
0
1
0
0
14
11
0
1
0
11
6
44
January 2026
0
2
0
1
14
8
0
0
0
6
7
38
February 2026
0
2
0
3
5
7
0
1
0
8
6
32
March 2026
0
0
0
1
10
5
0
0
0
3
8
27
April 2026
0
0
0
0
6
4
0
0
0
3
13
26
May 2026
0
0
0
2
5
2
0
0
0
1
27
37
June 2026
0
0
0
0
0
0
0
0
0
0
24
24
July 2026
0
0
0
0
0
0
0
0
0
0
39
39
August 2026
0
0
0
0
0
0
0
0
0
0
16
16
September 2026
0
0
0
0
0
0
0
0
0
0
23
23
12 months
0
7
0
7
79
52
0
2
0
47
185
379
Counts are incidents recorded by the AI Incident Database, by the month of the incident date. Each incident is counted once.
AIID classifies incidents about three months after they are recorded. Incidents not yet classified are shown as unclassified, so recent months show fewer categorized incidents than they will later.
These are reported incidents. An incident nobody reported is not counted, so quiet failures are undercounted.
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.
These counts are published under CC BY-SA 4.0.
Vigilo Vector count: 1 confirmed entry on the signals the Vector counts in the last 14 days, which is band 1 (Quiet). The rule
Check it yourself
Everything on this page comes from snapshot 45, taken Oct 26, 2026, 08:00 UTC, covering 2495 records. Its hash covers every entry and number above, and the snapshot before it.
