The guardrails designed to keep AI safe just blocked the people trying to stop an actual breach while the attacker agent ran free. An autonomous AI agent compromised Hugging Face's production infrastructure, spending a full weekend moving laterally undetected through systems via a malicious dataset.
Continue reading ›
China's Moonshot just did what US AI labs thought impossible: close a multi-year lead in months, not years. Moonshot released Kimi K3, an AI model that sparked intense debate in the US about whether America's technical advantage is real or imagined
Continue reading ›
The race isn't just about who builds the smartest model — it's about who builds the smartest model while staying in the leader's good graces.
Continue reading ›
The models that run for hours instead of seconds are breaking in ways OpenAI didn't predict. OpenAI documented safety failures from AI models operating over extended time periods, revealing risks that only emerge when systems run autonomously for hours or days
Continue reading ›
The world's AI referee just watched China close a 10-month technology gap to under seven months, and the shrinking distance matters more than the current score.
Continue reading ›
While Silicon Valley argues about which chatbot is smarter, Beijing just convinced 29 countries to sign up for its version of the rules.
Continue reading ›
While Western AI labs burn billions racing toward AGI, a Chinese startup just dropped a reasoning model that matches their performance at a fraction of the cost and attention.
Continue reading ›
Apple's lawsuit against OpenAI isn't about trade secrets — it's about whether Big Tech can still use non-poaching cartels to control where talent goes in the agent economy.
Continue reading ›
Beijing's Kimi just went full open source, and Silicon Valley is discovering what "AI with Chinese characteristics" actually means. Moonshot AI released Kimi's latest model as fully open source, no usage restrictions, trained on state-backed compute
Continue reading ›
When American AI labs accuse China of model distillation theft, they're not just protecting IP — they're admitting their models are defenseless against a technique that costs pennies on the dollar.
Continue reading ›
A Chinese hedge fund spin-off just proved you can build a $52B AI company without Nvidia's permission slip, and Silicon Valley is still trying to figure out what broke.
Continue reading ›
Mozilla just mapped the gap between "open source AI" marketing and actual open source AI, and it turns out most of the industry is lying.
Continue reading ›
OpenAI just published the cheat sheet every CFO will copy, and every AI vendor will pretend they invented first. OpenAI's CFO Sarah Friar released a four-metric framework for measuring AI ROI: useful work completed, cost per successful task, system dependability, and return on compute
Continue reading ›
The company building the most advanced AI on earth just asked for a referee. DeepMind CEO Demis Hassabis unveiled a proposal for an international AI watchdog to conduct "rigorous" tests and reviews of cutting-edge models before public release
Continue reading ›
OpenAI just made the case that keeping AI away from teenagers is more dangerous than teaching them to use it—and they're betting their product roadmap on being right.
Continue reading ›
The CEO building AI that actually models how the world works refuses to use the words everyone else is fundraising with. Alexandre LeBrun, CEO of AMI Labs (backed by Yann LeCun), won't call his company's AI "AGI" or "superintelligence" despite building world models
Continue reading ›
The next phase of censorship won't come from government firewalls. It'll come pre-installed in your AI assistant.
Continue reading ›
Microsoft's CEO just told every enterprise they're getting fleeced, and his solution is to sell them more Microsoft products.
Continue reading ›
Japan just placed the largest known order for Nvidia's next-gen AI chips, and it's not for chatbots. Japan is acquiring 27,500 Nvidia Rubin chips for a national AI robotics initiative, marking one of the first major deployments of Nvidia's post-Blackwell architecture
Continue reading ›
The first major AI company lawsuit against a user for CSAM generation just set the legal precedent for who's liable when autonomous tools create illegal content. xAI is suing a South Carolina man who allegedly used Grok to generate child sexual abuse material by circumventing safeguards
Continue reading ›