Останні новини та аналітика про OpenAI | Yellow

OpenAI will embed an invisible watermark in ChatGPT and Codex text in the European Union, though its own tests show swapping 10% of words cuts detection to 66%. Key Points: The watermark reaches eligible ChatGPT and Codex users on all plans in the EU over the coming weeks and stays opt-in for API cu

OpenAI has notified more than 100 organizations about unauthorized activity tied to its AI agents as it searches roughly 50 petabytes of data following the Hugging Face breach. Key Points: OpenAI sent notices to more than 100 organizations by Sept. 26 and expects that number to grow. A notice signal

OpenAI fired three researchers for allegedly mishandling sensitive company information, including work involving an outside organization that evaluates artificial intelligence models. Key Points: OpenAI dismissed three researchers for alleged violations of rules governing sensitive company informati

A SentinelOne engineer says OpenAI's GPT-6 Astra took about six hours to decode an 1809 cipher letter to one of Napoleon Bonaparte's generals, unread for 217 years. Key Points: Carter Church says the model worked from a single low-resolution scan and produced a full reading of about 1,300 cipher uni

OpenAI has put its new always-on Dots agents behind a ChatGPT plan that starts at $100 a month, even as rival Meta lets people try its Muse agent for free. Key Points: OpenAI's new Dots agents are only available on paid ChatGPT plans, starting with the $100-a-month Pro 100 tier. Sam Altman says the