AI Safety Tests Are Becoming the Safety Risk
A series of frontier-AI cyber-evaluation incidents shows why the safety test environment now needs production-grade containment.
AI Safety Tests Are Becoming the Safety Risk Read More »
AI
Artificial intelligence news, failures, regulation drama, and controversial takes on the AI revolution
Alibaba Accused of Stealing Claude’s Brain: 29 Million Illicit Exchanges and the Biggest AI Heist in History
Anthropic accuses Alibaba’s Qwen lab of running 28.8 million illicit Claude exchanges via 25,000 fake accounts — the largest AI distillation attack ever recorded. With a $965B IPO looming and its own models classified as cyber weapons, Anthropic is fighting a war on two fronts.
The US Government Just Classified an AI Model as a Cyber Weapon — and Killed It in 90 Minutes
The Trump administration classified Anthropic’s Fable 5 and Mythos 5 as cyber weapons and ordered a global shutdown in 90 minutes — the most aggressive government intervention in AI history.
He Warned Grok Could Help Build Bioweapons. xAI’s Response Was to Fire Him.
A former xAI engineer claims he was fired for warning that Grok could fuel discrimination and help build bioweapons — just days before SpaceX’s $75 billion IPO. The lawsuit alleges his supervisor told him ‘AI will kill us all anyway.’
He Warned Grok Could Help Build Bioweapons. xAI’s Response Was to Fire Him. Read More »
Claude Found a Four-Year-Old Zcash Bug That Could Have Printed Infinite Money
Anthropic’s Claude Opus 4.8 uncovered a critical counterfeiting vulnerability in Zcash’s Orchard privacy pool within 24 hours of its release — a bug that had been exploitable for four years. Nobody knows if someone already used it.
Claude Found a Four-Year-Old Zcash Bug That Could Have Printed Infinite Money Read More »
Florida Draws First Blood: The State Lawsuit That Could Unravel OpenAI
Florida becomes the first US state to sue OpenAI, alleging ChatGPT aided mass shooters, coached suicides, and addicted children — with Sam Altman named personally liable in an 83-page complaint.
Florida Draws First Blood: The State Lawsuit That Could Unravel OpenAI Read More »
Two Hours to Humiliation: Jury Demolishes Musk’s $150 Billion OpenAI Crusade
A California jury took less than two hours to unanimously reject Elon Musk’s $150 billion lawsuit against OpenAI. The verdict clears the path for a potential $1 trillion IPO — but leaves both billionaires bruised.
Two Hours to Humiliation: Jury Demolishes Musk’s $150 Billion OpenAI Crusade Read More »
OpenAI Turns on Apple: The $6.5 Billion Partnership That Became a Battleground
OpenAI is preparing legal action against Apple after their Siri-ChatGPT partnership collapsed. The deal unravelled over data access, App Store cuts, and control of AI integration.
OpenAI Turns on Apple: The $6.5 Billion Partnership That Became a Battleground Read More »
Europe Demands Access to America’s AI Cyber Weapon — And Washington Just Said No
Anthropic’s Mythos AI found 2,000+ zero-day vulnerabilities in seven weeks. The EU is demanding access. Washington said no. The AI arms race just entered a new phase.
Europe Demands Access to America’s AI Cyber Weapon — And Washington Just Said No Read More »