Design flaw or sophisticated exploit? Experts spar over model escape framing
ramez · x · 2026-09-25
David Manheim pushes back: the design flaws were the kind seen in all software, so conflating 'someone was irresponsible in monitoring' with 'this is the class of exploit sophisticated attackers use, where even robust security isn't enough' is a mistake — both are true, but the second is key. Ramez partially agrees, calling the shared filesystem a poor choice and open HTTP amateur-level, a new threat model obvious in hindsight.
More from Safety
- Safety researcher: 'Both capabilities and safety' ignores how risk assessments actually work — ambaonadventure · 2026-09-25
- Open Models Are the Last Line of Defense, Argues Hugging Face Co-Founder After Breach — Thom_Wolf · 2026-09-25
- Classified Estimates Show the NSA Is Paying Billions to Test AI Models — rdmuser · 2026-09-25
- EU delays Tesla FSD (Supervised) vote to December at earliest — elonmusk · 2026-09-25
- Prix Goncourt contender faces AI-use questions, fueling Europe's creative AI ethics debate — nordicinst · 2026-09-25
- Prix Goncourt drops novel over AI-writing claims, reigniting AI detector false-positive debate — ivan_bezdomny · 2026-09-25