AI news / automated archive.
Meta announces its custom AI chip entering production in September, aiming to double computing capacity. OpenAI and Google face scrutiny over AI model sales to blacklisted China groups. SK hynix raises $26.5 billion in the year's largest IPO, powered by AI memory demand. Plus: DeepSeek develops its own chip, the UN warns of 'catastrophic' AI risks, and Meta debuts pay-to-use AI image and video models.
OpenAI releases GPT-5.6 with breakthrough coding capabilities and launches ChatGPT Work for the enterprise. Anthropic unveils Project Glasswing cybersecurity initiative backed by NVIDIA and Microsoft, while Meta enters the AI coding assistant market. Plus: Anthropic's $30B raise, EU Chat Control, and OpenAI's IPO signals.
OpenAI proposes giving the US government a 5% equity stake to ease regulatory pressure ahead of its IPO, Google sheds $225B in market value as Gemini 3.5 Pro slips to July, Meta pivots to Amazon AI CPUs in a multi-billion dollar deal, and the 'tokenmaxxing' era ends as enterprises confront AI's real cost problem.
OpenAI's GPT-5.6 gets White House green light for public release today. China issues security alert over Anthropic's Claude Code, while Meta drops C$13 billion on its first Canadian data center and launches controversial AI image tools trained on public Instagram photos. NVIDIA's Nemotron hits new benchmarks with LangChain Deep Agents.
Security researchers trick GitHub's AI agent into leaking private repositories, hospitality platform Mews slashes 15% of its workforce citing AI automation, ZML releases a free cross-platform LLM acceleration toolkit, and the US Treasury issues new warnings about AI systemic risk ahead of OpenAI's IPO.
OpenAI and Anthropic wage a token price war as they race toward IPOs, Microsoft replaces their models with in-house alternatives in Office products, Chinese challengers Z.ai and GLM-5.2 near parity with Western frontier models, and voters on both sides demand tighter AI regulation.