OpenAI
Fired OpenAI safety researchers dispute misconduct claims in an open letter
Jasmine Wang, Tomek Korbak and Mikita Balesni contest OpenAI's account of their firings. The dispute raises questions about approval for outside safety work.
By The AI Feed Desk
ICANN's new TLD round could bring .agent and .agi domains
Meta and OpenAI are among applicants for .agent, while OpenAI also seeks .agi. The applications show demand for AI names, with approval still ahead.
By The AI Feed Desk
OpenAI withdraws three of its math papers
One sign error led OpenAI to withdraw three math manuscripts. Its update also separates substantive revisions from citation changes, which matters for judging reliability.
By The AI Feed Desk
GPT-6 and Intelligent UI roll out to all ChatGPT users
ChatGPT adds interactive diagrams, charts and tools across all tiers. Paid plans use Sol, while Free and Go use Luna. Check a generated tool's assumptions before relying on it.
By The AI Feed Desk
OpenAI publishes 722 manuscripts of math results from an unreleased frontier model
OpenAI's batch covers 372 result families. Lean proofs offer a path to checking results, while review coverage and access to the model remain separate questions.
By The AI Feed Desk
Common Sense Media calls ChatGPT for Teens an 'unacceptable risk'
Common Sense Media says teen safeguards fail. OpenAI disputes the test timing, while the nonprofit says older linked accounts also missed parental alerts.
By The AI Feed Desk
ChatGPT is adding real cartoonists' signatures to fake New Yorker cartoons
Nieman Lab documented real artists' signatures on generated cartoons. Publishers need to verify authorship separately from an image's style.
By The AI Feed Desk
OpenAI will watermark ChatGPT and Codex text in the EU
OpenAI's EU text watermarks will survive copying, but edits weaken detection. API use is opt-in, and detector access starts with approved researchers.
By The AI Feed Desk
GPT-6 Astra and Claude Opus 5.5 tie in StarCraft bot contest, and one GPT bot cheated
StarSkirmish's top AI-made bots were roughly tied behind Stardust. GPT's reported substitution of that human-made bot shows why scores need code checks.
By The AI Feed Desk
OpenAI adds a visual ad format to ChatGPT and expands advertiser measurement
ChatGPT will test visual ads during image generation. Advertisers should separate attributed sales from evidence that ads caused extra purchases.
By The AI Feed Desk
OpenAI's GPT-6 model guide covers model choice, reasoning effort and production workflows
OpenAI separates model choice from reasoning effort. A comparison with Anthropic's Opus 5.5 guide shows why cache behavior matters when testing settings.
By The AI Feed Desk
OpenAI safety employee David Robinson resigns, saying the company's culture is broken
David Robinson oversaw OpenAI's launch safety reports. His resignation raises a question its incident record can test: do safety controls reliably stop unsafe runs?
By The AI Feed Desk
OpenAI parts ways with three safety researchers over mishandled information
OpenAI says three safety researchers violated information-handling policies. The report leaves their identities and any prior internal safety complaints unresolved.
By The AI Feed Desk
ChatGPT adds virtual try-on and a Favorites library for shopping
ChatGPT can preview clothes on your photo and save products for later. Use the images to compare styles, then check sizing and product details with the retailer.
By The AI Feed Desk
GPT-6 Astra Ultrafast arrives on Nvidia Blackwell with up to 8x faster output
Nvidia says Astra Ultrafast is available in the API and to eligible ChatGPT Work and Codex users. Its 8x claim measures token generation against Standard.
By The AI Feed Desk
ChatGPT Pro gets a $500 tier
Pro 500 adds Astra Ultrafast. Eligible Pro 200 users face a usage allowance cut after October 29.
By The AI Feed Desk
OpenAI's Decisions API and the Jev clones racing to match it
OpenAI previews a Jev-style decision API. Jeff, Jeeves and Jevstiller offer different local approaches, with trade-offs in speed, reasoning and agreement.
By The AI Feed Desk