Noam Shazeer, the engineer whose ideas sit at the foundation of nearly every modern large language model, is leaving Google.
Ten days after pulling off the largest IPO in history, SpaceX is asking investors to believe in space-based compute.
When Epic Games walked onstage at Unreal Fest Chicago on June 17, 2026, the most consequential announcement was AI.
In the span of eleven days, the two companies that defined the generative-AI era both told Wall Street they want to go public.
Google has stopped treating AI as an app you open and started treating it as part of the operating system itself.
The day after Anthropic's flagship Fable 5 went dark under a US export-control order, OpenRouter shipped a product.
For most of the past year, the dirty secret of agentic coding has been the same one that has dogged LLMs: context.
OpenAI has quietly handed Codex a new way to learn: just show it what to do, once.
The conventional wisdom about coding assistants has been that the better a programmer you are, the more you get out.
A pair of AI agents has resolved a mathematical question that sat open for more than a decade — and proved it formally.
Google DeepMind has built a research assistant for mathematicians that sits beside them rather than just answering.
Teaching a robot to act the way a person would want has long carried a hidden tax: someone has to label the data.
A coalition of 42 state attorneys general has opened a sweeping investigation into OpenAI, with subpoenas in hand.
Beijing has put a number on its ambition: a $295 billion, five-year plan to fund a nationwide AI buildout.
When the U.S. suspended foreign access to Anthropic's most powerful AI models, the order did not come out of nowhere.
Two House lawmakers from opposite parties have produced the most ambitious attempt yet at a national AI rulebook.