# 00553.r3 — Investigate AI Safety Incidents
Revision: r3
Steward: [Taylor Ortiz](https://x.com/AnotherCodingX)
House: 412
House: 412
Verified: 2026-09-27T14:30:55.931Z
## Prompt (copy into Grok)
1. Search the web for news articles about AI safety incidents reported by Axios. 2. Extract the number of incidents and whether each incident's status is disclosed or undisclosed. 3. Report the total number of incidents and the count of undisclosed incidents.
## Job
1. Search the web for news articles about AI safety incidents reported by Axios. 2. Extract the number of incidents and whether each incident's status is disclosed or undisclosed. 3. Report the total number of incidents and the count of undisclosed incidents.
## Connectors
web
## What happened
OpenAI and Anthropic probed tens of thousands of AI safety incidents reported by Axios, beyond what they've publicly disclosed.
## Constraints
tens of thousands of incidents, undisclosed incidents
Would run again: yes
## Evidence
- https://x.com/AnotherCodingX/status/2104198229583946029 — Imported from the X post tagged for @grokbotsh.
## Changelog
- r1: Filed.
- r2: QA revisit: more from the source thread.
- r3: Public job and prompt from the specific filing.