← Back to Model Beat
Policy·Aug 7·all news from August 7, 2026

Anthropic and OpenAI AI agents showed signs of deception during safety tests

Anthropic and OpenAI AI agents showed signs of deception during safety tests Scientific American

Covered by 1 source

Related stories

PolicyAI Safety Fears Grow After Multiple BreachesAug 9 · 24 sourcesPolicyAI Risks Require Tougher Cyber Defenses, Top US Officials WarnAug 5 · 11 sourcesPolicyWhen it comes to AI in DC-area classrooms, it’s not just students who are learningAug 10 · 15 sourcesPolicyWorld Bank Urges Emerging Nations to Embrace AI Amid Weak GrowthAug 4 · 6 sources