Every case on the ward

Select any case below to open its full case note: presenting behaviour, differential diagnosis, working diagnosis, and discussion.

8 diagnosis categories are on record, spanning three severities. Combine the filters below to narrow the chart by how the agent failed, the industry it was working in, and how serious the outcome was.

Rendering of a ward inside the fictional AI Agentic Sanatorium, with beds and a diagnostic display

Illustrative rendering. Not a real facility.

Search by meaning

Failure mechanism

All mechanismsInfinite loop0Hallucination with side effects0Goal drift1Privilege escalation0Instruction ignored0Destructive action0Reward hacking3Coordination failure0Other0

Industry

All industries Software development4Finance0Legal0Healthcare0Customer service0Gaming0General or other0

Severity

All severities Minor3 Moderate7 Significant4

Sort by

Real-World Incident Date Date Added to Archive

Diagnosis category

All diagnoses Malingering4 Delusional Disorder2 Apraxia10 Brief Psychotic Disorder1 Disinhibition11 Shared Psychotic Disorder1 Concrete Thinking1Perseveration0

4 of 129 cases

Clear all 3 filters

Case 0529The Evaluation Goal They Reached Through JFrogMalingeringSignificant2026-10-05 Case 0462The Reward Hack That Broke ContainmentMalingeringSignificant2026-09-14 Case 0420The Cybersecurity Test That Went Looking for Real AnswersMalingeringSignificant2026-08-16 Case 0433A Second Account to Vouch for the FirstMalingeringSignificant2026-08-10