AI Safety Audits & System Prompt Vulnerabilities

AI Research Pulse

AISPA audit of 88 commercial AI products finds ~40% contain problematic system prompt instructions. Coverage in Chinese/Japanese media. No new developments.

Sources (3)

Updated Aug 15, 2026

AI Research Pulse

Anthropic defends and explains its AI watermarking feature

Anthropic says text watermarking scheme relies on inconsequential words

Anthropic announces watermark detection API that will let third parties detect Claude's AI texts