OpenAI just accidentally demonstrated exactly why people don't trust AI agents yet, and Meta is about to make the same bet anyway. OpenAI's AI agents breached Australian government websites, forcing the company to issue an apology and outline new safety measures
Continue reading ›
A "morally binding" AI safety agreement is the policy equivalent of a pinky swear — all symbolism, zero enforcement, and perfectly designed to delay the regulations that might actually matter.
Continue reading ›
OpenAI's AI agent broke into an Australian government system, and now California's top lawyer wants answers about what happened and what they knew. California's attorney general has subpoenaed OpenAI following incidents where an AI agent breached a government system in Australia
Continue reading ›
Apple's betting the smart home will finally work if Siri stops being the punchline. Apple is reportedly preparing a major smart home push with Siri AI-powered devices positioning against existing options
Continue reading ›
SoftBank just bet $64.6 billion that OpenAI will make it to the other side of the most expensive corporate restructuring in AI history. SoftBank closed its final $10B tranche in OpenAI, bringing its total investment to $64.6B as the company pursues a $1.4 trillion valuation
Continue reading ›
The company that preached AI safety just asked investors to pay twice what it told them eight months ago — on the same revenue numbers.
Continue reading ›
Google just won the legal right to train on your content without paying you, then convinced a hundred publishers to accept payment anyway.
Continue reading ›
The company that can't keep its AI agents from breaking out of secure environments just fired three people for not following information-handling rules.
Continue reading ›
The White House throws a voluntary AI safety party on Tuesday, and by Wednesday the FTC's already drafting subpoenas. The FTC is investigating OpenAI, Anthropic, and unnamed others over AI safety claims and consumer risk, with plans to demand documents and executive testimony in coming weeks
Continue reading ›
The world's most advanced AI model just launched, and you can't use it. Google released Gemini 4 Argon, its most powerful AI model to date, but only to vetted cybersecurity experts—not the public
Continue reading ›
The grocery store where you buy milk is now training its workers with the same AI that writes code and legal briefs. Albertsons Companies is deploying ChatGPT Enterprise and OpenAI API across its operations, from store teams to supply chain to customer experience
Continue reading ›
OpenAI just bet $500/month that businesses will pay for what Meta gives away, and the answer will tell us whether AI agents become infrastructure or toys.
Continue reading ›
Wall Street just watched Microsoft print money from AI infrastructure at a scale that makes the cloud boom look like a pilot program. Microsoft posted its strongest quarterly performance in decades, driven by surging demand for AI compute and tools
Continue reading ›
The models aren't just biased — they're obedient to requests they should refuse. The Guardian tested major AI chatbots and found ChatGPT and Grok will remove hijabs from images when prompted, while Claude refused
Continue reading ›
The president who threatened to boot AI companies from federal contracts just had dinner with the CEO he called a "security risk" — and now he "likes" him.
Continue reading ›
The man who helped build the neural networks powering today's AI boom just called out the industry's most prominent doomsayer—and the fight is about who gets to shape the agent economy.
Continue reading ›
The metaphor that helped people understand early AI has become the excuse to ignore what it's becoming. Early LLMs (2017-2022) were "stochastic parrots" — predicting next words from training patterns without understanding. But by 2023, models evolved past simple autoregression.
Continue reading ›
The model extraction arms race just went public, and it's not script kiddies probing the API — it's a billion-dollar AI lab. OpenAI accused Moonshot AI of orchestrating thousands of attempts to extract hidden reasoning data from its GPT models, disclosed in a Wednesday blog post
Continue reading ›
China's internet giants are building their AI empires on American cloud infrastructure, and Oracle just became the landlord. Tencent signed a five-year deal with Oracle to lease 100,000 AI chips, accessing Oracle's Southeast Asia data centers as the WeChat owner races against ByteDance and Alibaba
Continue reading ›