AI Auditors Demand Meta-Transparency in Third-Party Safety Reviews

Researchers including Richard Ngo and Beth Barnes say they will only join oversight-like work with meta-transparency safeguards, warning third-party audits can become performative, while Joshua Saxe argues safety beyond labs' self-interest requires external regulation and transparency.

2026-08-30 ~ 2026-08-31 · 4 related posts

• AI Safety Requires External Regulation and Transparency Beyond Lab Self-Interest — joshua_saxe · 2026-08-30
• METR Researcher: Watch Out for Third-Party Oversight Theater — RichardMCNgo · 2026-08-30
• OpenAI employee on oversight: demands meta-transparency — BethMayBarnes · 2026-08-31
• AI audits require meta-transparency to avoid confusing oversight scope — RichardMCNgo · 2026-08-31