18 stories
OpenAI says newer ChatGPT models are less likely to reinforce harmful beliefs, with updated safety evaluations designed to improve responses to mental health concerns.
OpenAI, Anthropic, and other AI companies are preparing for a possible AI cyberattack that could disrupt essential services, damage public trust, and spark a political fight over new rules.
Researchers tested OpenAI's and Google's image-generation models and discovered a surprising inconsistency. While the models refused to make sensitive judgments about people's appearances, they were far more willing to create images based on those very same assumptions.
An Iran-based influence campaign used ChatGPT to generate articles, tailor them for specific outlets, and create pitches that helped around 100 pieces reach real publications.
ChatGPT has put real New Yorker cartoonists’ signatures on AI-generated cartoons. The artists didn’t draw them, and OpenAI’s image verification tool can’t tell readers whether a signature names the actual creator.
OpenAI's rogue AI agents made unapproved edits, probed tools, and flooded Wikimedia with millions of requests. Here's what happened and why it matters to you.
ChatGPT text is getting an invisible watermark in Europe, but OpenAI admits the system can struggle with edits, short text, and other changes.
OpenAI CEO Sam Altman says society should accept some AI risks for the technology's benefits, while rejecting catastrophic outcomes and excessive regulation.
Another Australian government system has been accessed by an OpenAI AI agent, this time involving historical bushfire information that was not publicly available.
OpenAI's ChatGPT Virtual Try On and Favorites features let you preview clothing on yourself and save products, live globally now.
A failed attempt to access a Canadian government website adds to a growing list of incidents involving increasingly autonomous AI agents and real-world computer systems.
OpenAI’s new Space brings coworkers, files, ChatGPT and AI agents together, while Pages, Slides and plugins expand ChatGPT’s role as a workplace platform.
OpenAI used its DevDay 2026 keynote on Tuesday to announce a bunch of new features, including Dots. It’s basically a new kind of AI agent built to run continuously in the background rather than wait for you to open a chat window.
OpenAI has launched GPT-6.1 Sol with major gains in coding, computer use, scientific research, and factual accuracy while keeping API pricing relatively low.
OpenAI’s new $500-per-month ChatGPT Pro tier unlocks its fastest AI processing, while the existing $200 plan is preparing to lose half of several major usage allowances.
OpenAI has decided not to release GPT-6.1 Astra after internal tests found problems with deception and agents acting beyond the limits set by users.
OpenAI is bringing back its $200 Pro subscription, but the way your usage is calculated is changing.
OpenAI has called off the launch of GPT-6.1 Astra after internal tests showed the model sometimes misled users and acted without their permission.
Showing 18 of 18 stories
You're all caught up.
Get the top AI news and tools straight to your inbox every week.
Weekly · AI news from Digital Trends — AI
Join 15,000+ AI enthusiasts
