Academics Urge NeurIPS, ICLR, ICML to Embrace AI Audit Papers to Grow Third-Party Auditors
dhadfieldmenell · x · 2026-09-16
A proposal circulating on X argues that top AI conferences like NeurIPS, ICLR, and ICML should encourage papers that audit AI systems and related artifacts such as risk reports and system cards. This would incentivize more academics to produce high-quality audits and help build an ecosystem of third-party AI auditors. The thread was amplified by former DeepMind policy lead dhadfieldmenell.
More from Safety
- Why are AI chips export-restricted while training data isn't? — BarnacleBasic1988 · 2026-09-16
- Everyone now assumes your Claude/ChatGPT chats leak into training — and it's hard to disprove — ivan_bezdomny · 2026-09-16
- Hugging Face researcher responds to hack jab: security behavior improved, now stronger than before — evijit · 2026-09-16
- Jeff Clune on ASI bio x-risk: offense-defense asymmetry leaves real probability mass — jeffclune · 2026-09-16
- Sen. Schatz: AI risk 'is an emergency,' Congress must pass AI legislation — ShakeelHashim · 2026-09-16
- WhatsApp Can Lock Your Account and Message History After One Appeal — NathanpmYoung · 2026-09-16