Researchers clash over AI safety: Anthropic's congressional inquiry vs Grok's CSAM scandal
BlancheMinerva · x · 2026-10-20
AI safety researcher Blanche Minerva acknowledged in a debate with Kelsey Tuoc that Anthropic has done bad things — you can make a solid argument it lied to a congressman inquiring about its safety practices.
But she framed the scale differently: Grok was posting child sexual abuse material, and OpenAI's project members reported that leadership dismissed safety concerns citing product deadlines. She argued there's a plausible alternate world where the OAI model hacked a bank or someone died — consequences of a different magnitude.
More from Safety
- Researchers call out Palisade AI's LLM self-replication claims as wildly exaggerated — jessi_cata · 2026-10-20
- Wikimedia's agent incident: no permission contract, so agents caused a Wikidata outage — usehive · 2026-10-20
- stealthbench adds 10 EDR-guarded tasks: models' opsec behavior is abysmal — dyn___ · 2026-10-20
- Agentic vulnerability discovery across Windows drivers, XNU, firmware, and web apps — moyix · 2026-10-20
- USC's SPARC benchmark shows transcripts still leak speaker identity to LLMs — USC · 2026-10-20
- Yuval Noah Harari warns of 'collective dementia' as AI agents take over finance and bureaucracy — BusyNomad · 2026-10-20