OpenAI is rolling out DALL·E 3 to ChatGPT Plus and Enterprise users. The release is backed by a multi-layer safety mitigation stack designed to reduce potential harms as access to the image generation…
OpenAI has launched a Safety Bug Bounty program inviting external researchers to find and report security flaws and misuse scenarios in its AI models, with rewards for valid findings. The program targ…
This article explores the OpenAI mission: ensuring artificial general intelligence benefits all of humanity. It explains how the mission guides daily decisions, from prioritizing safety testing and al…
OpenAI's Model Spec is a public framework defining how its AI models should behave—covering instruction-following, refusals of harmful requests, honesty, and limits on misuse. The article explains how…
AGI refers to AI systems matching or exceeding human abilities across many tasks. Because such technology could transform healthcare, education, work, and science, responsible development is essential…
OpenAI unveils a new multimodal moderation model built on GPT-4o, offering improved detection of harmful text and images. This upgrade to the Moderation API provides developers with a stronger foundat…
This article explains the safety testing behind OpenAI's o3-mini model in plain language. It covers three layers of scrutiny: internal safety evaluations checking whether the model refuses harmful req…
OpenAI introduced Trusted Contact, an optional ChatGPT safety feature that alerts a person you designate if a conversation shows serious signs of self-harm risk. The article explains how to enable it,…
This article explains in plain language how frontier AI labs' revised preparedness frameworks work: labs run capability evaluations to test whether their most powerful models could enable severe harm …
This report examines the safety and capabilities of OpenAI's o3 and o4-mini models. Both models combine cutting-edge reasoning with full tool access, including web browsing, Python, image and file ana…
OpenAI and Microsoft have signed a new Memorandum of Understanding reinforcing one of the most important alliances in AI. The MOU reaffirms Microsoft Azure as central to OpenAI's model training and de…
OpenAI has proposed creating an international organization dedicated to keeping young users safe in the age of AI. The proposal calls for stronger safeguards, shared global standards, and expanded acc…
OpenAI has announced a public policy agenda urging cooperation between governments and the AI industry. It covers four priorities: building safe and reliable AI systems with independent evaluation, pr…
OpenAI released gpt-oss-safeguard, a pair of open-weight reasoning models (120b and 20b parameters) designed for safety classification. Unlike traditional moderation tools, developers write policies i…
OpenAI's Teen Safety Blueprint is a framework for building AI that teens can use safely. It covers age-appropriate design, default protections like parental controls, privacy balance, and expert input…