model evaluation

AI News Report — Accurate. Independent. AI-Focused. · https://www.ainewsreport.org.njangi.app/tag/model-evaluation

AI Safety & Ethics Investigation Critical risk

UK and US Safety Institutes Find Kimi K3 Safeguards Failed to Block Offensive Cyber Attempts Before Open-Weight Release

Pre-release evaluation of an open-weight model has a property that evaluation of an API-served model does not: once the weights are out, the findings describe s...

25 Jul 2026 3 min read United Kingdom United States Global

Global Governance Analysis Medium risk

The Safety Institutes Network: What Coordination Between States Can and Cannot Achieve

A network of national institutes solves a coordination problem and leaves an authority problem untouched. Both are worth being precise about.

10 Jul 2026 2 min read Global United Kingdom United States