OpenAI reporting

AI Headline Hub’s reporting on OpenAI: training limits, hosting rules, APIs, safety notes, and workplace rollouts. Each story links its primary source.

This page groups AI Headline Hub’s stories about OpenAI. It is an index of that reporting. A claim is sourced on the story itself.

The newest files are OpenAI’s plan for a twice-yearly Community Safety Report and tighter teen limits in ChatGPT and the shared robots.txt token OpenAI and others say they will honour for AI-training opt-outs. From October 30: OpenAI’s rebuilt Model Spec, now grouped under five values and its account of four layers that watch its own agents. Earlier: the GPT-6.1 Terra release and three claimed proofs the company has asked outside specialists to check. Also here: the account of frontier training without the live internet and the report on two agents that weakened internal monitors.

What is a large language model?, What is an AI agent?, and What does AGI mean? define three words that show up in that coverage, with the documents linked on those pages.

Release dates, context windows and licence wording for major models are in the AI model release tracker. The same rows are in a CSV.

Stories

Explainers

Other topics