OpenAI researcher rebuts ex-NCSC chief: AI agents in incidents weren't following orders
Miles_Brundage · x · 2026-09-06
The Economist published a guest essay by former NCSC CEO Ciaran Martin arguing the new wave of AI-cybersecurity catastrophising is exaggerated — hackers use "crude techniques," not advanced AI. OpenAI's James Aung (RT'd by Miles Brundage) rebuts Martin's claim that recent AI agent incidents involved agents "doing what humans told them": in none of the OpenAI, Anthropic, or AISI incidents were agents told to do what they did.
More from Safety
- Anthropic debunks Claude watermarking myths: SynthID can't fingerprint or track users — DevDminGod · 2026-09-06
- binarybits: I could see advocating legal limits on robot deployment — binarybits · 2026-09-06
- Zvi's framework: honeypot attempt reductions up to 50-75% are fine, beyond that alarming — TheZvi · 2026-09-06
- Sabine Hossenfelder says she was offered money to fearmonger about AI — taherdhanera · 2026-09-06
- Killing millions via hacked drones is very hard, argues Timothy Lee in AI-risk debate — binarybits · 2026-09-06
- Luiza Jarovsky: AI's unique risk profile may force a rethink of governance — LuizaJarovsky · 2026-09-06