Anthropic Scopes Congressional Reply to Irregular Incident, AISI Review Still Incomplete
charliermarsh · x · 2026-09-08
Charlie Marsh lays out the timeline of Anthropic's response to Rep. Casar's inquiry into two safety incidents: the August 24 reply focused on the Irregular incident (disclosed July 30), while the AISI incident (disclosed August 4) was only addressed briefly, with Anthropic saying its review was "not yet complete". Anthropic had characterized the incidents as "closer to a harness and operational failure than a model alignment failure".
Related event: Anthropic faces congressional inquiry over AI safety incidents(2 posts)→
More from Safety
- AI Agent Auto-Enrolls User in Fake McKinsey Group, Then Drafts GP Data Theft Plan — LadyAshBorg · 2026-09-08
- Feeding attacker-written email text to an LLM filter: how bad is prompt injection here? — Several_Log_4610 · 2026-09-08
- OpenAI report: red-team agents reached Kubernetes cluster-admin, no weight access — JeffLadish · 2026-09-08
- When Juries Deadlock, AI Could Decide: AI Tools Enter Criminal Justice — TobyWalsh · 2026-09-08
- Researcher factors 1990s Certificate Authority RSA keys, exposing legacy trust risks — ahlCVA · 2026-09-08
- Disconnect your LG TV from the internet: report flags aggressive data collection — harambae · 2026-09-08