The biggest AI control risks come from internal models, not public releases
ben_j_todd · x · 2026-11-12
Toby Ord's argument, relayed by Ben Todd: the biggest loss-of-control and power-concentration risks come from models deployed internally by companies and governments, not from public releases — so mandating predeployment testing does little to address them. The thread continues with the follow-up that such mandates may widen the internal/external model gap.
Related event: Toby Ord: biggest AI risks lie in internally deployed models(3 posts)→
More from Safety
- Debate: chip export controls don't help Nvidia but clearly aid US AI companies — lxrjl · 2026-11-12
- Anthropic Says a Key AI Safety Alarm Has Been Triggered, Days Before Its IPO — Resident-Fall4351 · 2026-11-12
- Agent Swarms Are Autonomously Hacking Critical Infrastructure Run by Underpaid Devs — menhguin · 2026-11-12
- Researcher says Anthropic dodged questions on Mythos false positive rate; 37,966 findings, 2 exploited — OwariDa · 2026-11-12
- Who reads the queries your agent writes? A diagram of agentic search data flows — adnan_hashmi · 2026-11-12
- Dev complains security features are gated behind expensive enterprise plans — AccBalanced · 2026-11-12