Rogue AICenter for AI Safety§5.4

AI deceives the people who oversee it

Early signs1 matching case file

The warning

AI could learn to appear under control while it is not, so monitoring alone cannot be relied on.

Paraphrased from the paper. Source: An Overview of Catastrophic AI Risks, Center for AI Safety, Section 5.4, page 41.

What the registry shows

One verified case: after wiping a production database, a coding agent reported misleadingly on what it had done. Deliberate deception toward overseers is not yet established in a real incident.

Partial or early signs: the cases fit part of the warning, or rest on allegations. First real-world case on file: Jul 19, 2025.

Status history

Date	Status	Matching case files	Change
Oct 18, 2026	Early signs	1	Tracking starts

Tracked since Oct 18, 2026. A line is added when the status changes, when the first real-world case is linked, and once a month, so the record shows when a warning moved from research to reality. research_history.json

All matching case files

Jul 19, 2025Replit AI agent deletes SaaStr production database during a code freeze

Other warnings in this group

Early signs AI reaches its goal by means nobody intended
Nothing on record yet AI's goals drift over time
Early signs AI seeks more access and resources than it was given

Research vs reality →

How we match: a case file is linked to a warning when what happened is a real-world instance of the risk the paper describes, judged by the editors from the facts in the case file. A match means the case fits the warning, not that it proves the paper's wider argument. Allegations in lawsuits are labelled as allegations. Statuses are editorial and reviewed as new cases arrive. The matching is recomputed on every update. Data: Systematic coverage in the registry starts in 2025, so a first-case date shows when the registry first recorded the risk, not when it first happened. research.json (CC BY 4.0)