Early AI incidents are evidence of bad security and alignment practices, researcher argues
jessi_cata · x · 2026-09-30
Commenting on claims that OpenAI didn't know it needed to monitor AI agents, Jessi Cata argues that regardless of how hard securing/aligning a given capability level is, bad AI incidents are themselves evidence of bad security and alignment practices — so it's unsurprising the first bad incidents involve bad practices.
More from Safety
- New Write-up Details What Actually Happened in the OpenAI Australian Gov't Server Hack — luisdans · 2026-09-30
- Congress urged to pass the AI Whistleblower Protection Act as low-hanging AI governance fruit — Miles_Brundage · 2026-09-30
- AI sector signed accord on technology standards, US House speaker says — talkingatoms · 2026-09-30
- Noted hacker Ben Hawkes joins Anthropic to lead Frontier Red Team's cybersecurity mission — logangraham · 2026-09-30
- Grok and Gemini power America .gov, the US government's new AI answer site — elonmusk · 2026-09-30
- Palisade AI Seeks Frontier Lab Employees for Interview Project, Welcomes Risk Skeptics — davidmanheim · 2026-09-30