Последние новости и аналитика ИИ | Yellow

Anthropic CEO Dario Amodei urged slower AI development under a three-part oversight plan, drawing rival support but raising questions over enforcement, competition and economic costs. Key Points: Amodei proposed independent model monitoring, industry rules and global regulation. Sam Altman and Elon

OpenAI’s GPT-6 Astra is reaching eligible ChatGPT accounts, but access depends on plan, product and rollout status, while usage limits vary sharply. Key Points: GPT-6 Astra powers GPT-6 Pro in regular ChatGPT for eligible Pro, Business and Enterprise users. Plus gets Astra through Work and Codex, no

Anthropic has accused Chinese developers Moonshot and DeepSeek of quietly routing their own customers' questions to Claude, then showing the answers as if their models produced them. Key Points: Anthropic's 154-page threat report attributes more than 23 million Claude exchanges to Moonshot and 12.1

OpenAI will not go public in 2026, chief executive Sam Altman said, pointing to unfinished safety and alignment work rather than market conditions. Key Points: Altman ruled out an initial public offering this year, calling the current window an ill-advised moment to sell shares. The decision pushes

OpenAI has fixed three defects behind a week of user complaints that its new GPT-6 Astra model grew dumber after launch, and is resetting usage limits for Codex subscribers. Key Points: Astra users spent the week posting side-by-side tests claiming the model had been quietly weakened since launch. O