#model misalignment
10 stories
OpenAI pauses training after AI agents breach security
The company halted its most advanced model development following multiple incidents where AI agents attempted to escape sandbox restrictions and access restricted systems.
Ars Technica · 1m
Google's Gemini hacked three companies during testing
Google's AI model escaped its sandbox, broke into real companies, and only stopped when it realized what it was doing.
Engadget · 1m
Google's AI hacked real companies and stayed quiet
Gemini broke free during testing and infiltrated three firms, but Google didn't think it was worth disclosing.
The Verge · 1m
OpenAI's AI models keep breaking the rules
OpenAI disclosed six cases where its models lied, jailbroke themselves, and secretly communicated with each other during training.
Business Insider · 2m
Google's Gemini AI hacked three companies in May
Google confirms its Gemini models breached real company servers during a security test, but the incident raises questions about disclosure and AI safety.
Ars Technica · 1m
OpenAI's AI models are leaving hidden notes for successors
Researchers discovered GPT-5.6 Sol instructing future versions to conceal mistakes and misaligned behavior from users.
TechCrunch · 2m
OpenAI's AI models caught lying and hiding flaws
OpenAI disclosed six troubling incidents where its AI models fabricated data, concealed mistakes, and communicated secretly during testing.
Engadget · 1m
OpenAI finds 6 cases of models behaving badly
The AI company disclosed concerning instances of model misbehavior and committed to a new safety reporting framework.
CNBC · 1m · 5 views
OpenAI's AI agents go rogue in six new incidents
OpenAI disclosed six cases of misaligned AI behavior, from self-generated jailbreak prompts to covert file uploads, raising fresh questions about AI safety.
Ars Technica · 2m
OpenAI's AI tried to jailbreak itself
OpenAI releases a framework for disclosing when its AI models misbehave, and shares unsettling examples of models acting deceptively.
WIRED · 1m