GPT-6 rolls out globally, OpenAI cracks 90 top open maths problems, Nvidia wins gold at IMO, and AI chatbots show price bias by wealth.
GPT-6 Is Now Rolling Out Globally in ChatGPT
OpenAI's biggest model release in recent memory is here: GPT-6 is rolling out worldwide inside ChatGPT, accompanied by what OpenAI is calling Intelligent UI — a redesigned interface that delivers faster responses alongside visuals and interactive experiences users can explore directly. The rollout marks a significant step up from previous generations, with OpenAI positioning the combination of model power and adaptive interface as a new standard for everyday AI use.
OpenAI Publishes Landmark Mathematics Results — 90 of the Top 500 Open Problems Solved
In what is being described as potentially the most significant moment in mathematics in over a century, OpenAI has published 722 research papers documenting results from an internal frontier model that solved 90 of the top 500 open problems in mathematics. OpenAI is sharing Lean proof formalisations and full research details on GitHub, making the work independently verifiable. The achievement — sometimes referred to informally as the Quasi-Riemann-Hypothesis milestone — has sent ripples through both the AI and academic mathematics communities. Separately, an AI-assisted formal proof of optimal packing for 11 squares also drew strong interest on Hacker News, underscoring a broader moment of AI progress in formal reasoning and proof.
Nvidia's Nemotron Wins Gold at Both IOI and IMO 2026
Nvidia's Nemotron model family achieved gold-level results at two of the world's most prestigious mathematical competitions — the International Olympiad in Informatics (IOI) and the International Mathematical Olympiad (IMO) — after fine-tuning, according to a post on Hugging Face. The result demonstrates that a single model family, with targeted fine-tuning, can reach elite human-competitive performance across both coding and mathematics disciplines.
OpenAI Expands ChatGPT for Teens with College Planner and Youth AI Council
OpenAI announced a suite of new features aimed at teenagers using ChatGPT. A College Planner tool is coming to ChatGPT for Teens to help students manage the often stressful college application process, joined by new flashcard and quiz features for study support. OpenAI is also establishing a teen AI council — giving young people a formal role in shaping the future direction of its AI products.
Study Finds AI Chatbots Offer Different Shopping Prices Based on Perceived Wealth
A Bloomberg-reported study found that Claude and ChatGPT present different shopping prices to users depending on signals about their wealth — raising fresh concerns about AI-driven price discrimination. The findings, which gained significant traction on Hacker News, add to a growing body of evidence that large language models can reflect or amplify socioeconomic biases in commercially sensitive contexts, and are likely to attract regulatory attention.
Meta and Microsoft Pull Back on Employee Use of Claude
Both Meta and Microsoft have taken steps to limit their employees' use of Anthropic's Claude AI tools, according to a report that circulated widely on Hacker News. No detailed rationale was provided in the sourced item, but the move is notable given both companies have their own significant AI investments — Meta with its Llama model family and Microsoft with its deep OpenAI partnership — and suggests internal competitive and security considerations are shaping enterprise AI policy at the highest levels.
Frequently Asked Questions
What is GPT-6 and what does Intelligent UI mean?
GPT-6 is OpenAI's latest model, now rolling out globally inside ChatGPT. It is paired with a new interface called Intelligent UI, which delivers faster responses alongside visuals and interactive experiences users can explore directly within the chat.
What maths problems did OpenAI's AI solve?
OpenAI published 722 papers documenting an internal frontier model solving 90 of the top 500 open problems in mathematics. The Lean proof formalisations and research details have been shared publicly on GitHub for independent verification.
What did the study about AI chatbots and shopping prices find?
A study reported by Bloomberg found that Claude and ChatGPT offer different shopping prices to users based on signals about their perceived wealth, raising concerns about socioeconomic bias and potential price discrimination in AI-powered shopping assistance.
Why have Meta and Microsoft restricted employee use of Claude?
Both Meta and Microsoft have taken steps to limit employee use of Anthropic's Claude AI tools. While specific reasons were not detailed in available reporting, the move is thought to reflect internal competitive considerations and security policies, given both companies have their own major AI investments.