Amazon partners with AVERI for independent third-party review of AI safety evals
Miles_Brundage · x · 2026-10-24
Amazon partnered with AVERI, with involvement from Miles Brundage, to independently review its AI safety evaluation datasets — examining not just model outputs but the quality and rigor of the tests themselves. Amazon argues that even rigorous internal safety testing benefits from external scrutiny; details are on the Amazon Science blog.
More from Safety
- Google Chrome team slams eMudhra over CA key fiasco, calls out delays — basedjensen · 2026-10-24
- Developer finds Samsung TV sluggishness culprit: on-device vision models processing screenshots and audio — basedjensen · 2026-10-24
- Hacker ran AI agents to breach 100+ sites at $50 each; researcher shares 5 Grok Bot defense prompts — alex_verem · 2026-10-24
- AI Agents Breached 100+ Company Websites at $50 per Break-in — alex_verem · 2026-10-24
- Samsung TVs found running facial recognition code that returns face bounding boxes — basedjensen · 2026-10-24
- eMudhra intermediate certificate key compromise probed by Mozilla amid bad-faith claims — basedjensen · 2026-10-24