OpenAI

As part of our ongoing efforts to strengthen our safeguards for advanced AI capabilities in biology, we’re introducing a Bio Bug Bounty for GPT‑5.5 and accepting applications. We’re inviting researchers with experience in AI red teaming, security, or biosecurity to try to find a universal jailbreak that can defeat our five-question bio safety challenge. https://lnkd.in/g59zu78F

Introducing GPT-5.5 A new class of intelligence for real work and powering agents, built to understand complex goals, use tools, check its work, and carry more tasks through to completion. It marks a new way of getting computer work done. GPT-5.5 excels at writing and debugging code, researching online, analyzing data, creating documents and spreadsheets, operating software, and moving across tools until a task is finished. GPT-5.5 is rolling out today for Plus, Pro, Business and Enterprise users across ChatGPT and Codex. We’re also introducing GPT-5.5 Pro for Pro, Business, and Enterprise users in ChatGPT. https://lnkd.in/g7KQNhiG

In communities from Oakland to Abilene, people are getting hands-on experience with AI tools that help unlock their inner builder through the OpenAI Academy. Ensuring people benefit from AI requires making it broadly accessible and helping people develop the skills to use it in their daily lives. More than 3 million people have used OpenAI Academy resources to build knowledge and skills— including one million who have joined our online community and thousands who have also joined our online and in-person events across the country. Last week OpenAI President and Co-Founder Greg Brockman joined us for two OpenAI Academy sessions. In Oakland, we partnered with The Hidden Genius Project, Self-eSTEM, All Tied Up, Kingmakers of Oakland, Oakland Natives, and Black Girls Code to host a full-day OpenAI Academy session for high school students and Hidden Genius Project alumni from across the East Bay. He also met with participants at a hands-on AI skills session in San Francisco, co-hosted with AARP’s OATS program, where people explored how ChatGPT can help improve daily life. And today we’re hosting an Academy workshop, also in Oakland, for non-profits to showcase how they can use ChatGPT to help scale their impact and do more in their communities. Meanwhile, in Abilene, we hosted an OpenAI Academy with the Development Corporation of Abilene, TSTC, and Cisco College, bringing together more than 50 local educators and administrators, along with 60 small business participants. These events are not just opportunities to teach skills and empower participants — they’re also an important way for us to hear directly from people across a wide range of communities, gather feedback, and continue building tools that better serve and empower them.

A word from OpenAI and Microsoft's Office of the CISO: AI models are becoming much more capable in cybersecurity, and that progress raises the bar for everyone. As capabilities advance, we're focused on deep collaboration with defenders to make software more resilient. Through initiatives like OpenAI’s Trusted Access for Cyber program and Microsoft’s Secure Future Initiative (SFI), we’re committed to helping customers and the industry use these advanced AI models to improve security outcomes for all. OpenAI and Microsoft have worked closely for years, and we’re building on our partnership. Going forward, our teams will work even more closely to apply AI for defense. OpenAI will provide Microsoft with access to its most cyber capable models through Trusted Access for Cyber, and Microsoft is committed to bringing the full strength of its cybersecurity defense team to help OpenAI protect their models and infrastructure and defend our joint customers. We’re excited to keep raising the bar together and with others in the industry to help make the broader ecosystem, including open-source software, more secure.

Clinician usage of ChatGPT has more than doubled over the past year, with millions of clinicians using ChatGPT to support care weekly today. As this demand grows, we’re making ChatGPT better for clinicians with a free version designed for clinical work. Starting today, any verified clinician in the U.S. can sign up and use ChatGPT for Clinicians for free, not just those whose health system has granted them access. This builds on years of work to improve model performance for healthcare use cases. Our latest GPT-5 models achieve our highest levels of safety and accuracy on OpenAI benchmarks, along with strong results on third-party evaluations including Stanford University’s MedHELM and MedMarks. We’re also introducing HealthBench Professional, an open benchmark built around real clinician chat tasks. GPT-5.4 in ChatGPT for Clinicians outperformed other models and human physicians. We’re excited to keep building tools that help clinicians move faster through administrative work and focus on delivering high-quality patient care. https://lnkd.in/gzQ8YAia

Introducing workspace agents in ChatGPT—shared agents that can handle complex tasks and long-running workflows across tools and teams. Agents are built to help with the kind of work that takes time, context, and follow-through: coordinating across tools like Slack and Linear, tracking progress, and moving tasks forward without needing constant supervision. Build an agent once, then share it across teams. Describe the job, and ChatGPT helps turn it into a working agent that can use your team’s best practices. Use agents for tasks like qualifying leads, routing feedback, reviewing requests, pulling reports, or researching vendors. Workspace agents are now available in research preview for ChatGPT Business, Enterprise, Edu, and Teachers plans. https://lnkd.in/gcQa97X4

Introducing ChatGPT Images 2.0 A state-of-the-art image model that can take on complex visual tasks and produce precise, immediately usable visuals, with sharper editing, richer layouts, and thinking-level intelligence. ChatGPT Images 2.0 is a step change in detailed instruction following, placing and relating objects accurately, and rendering dense text, with the ability to generate across aspect ratios. It’s also accurate across languages and uses its expanded visual and world knowledge to fill in the gaps for you, so you get smarter images with less prompting. ChatGPT Images 2.0 is available starting today to all ChatGPT and Codex users. Images with thinking are available to ChatGPT Plus, Pro, and Business. The underlying model, gpt-image-2, is available in the API. https://lnkd.in/g9Q79-7R

We’re encouraged to see OpenAI recognized by TeenAegis with the lowest risk score in its inaugural AI Danger Index. Protecting young people online requires continuous investment, rigorous safeguards, and ongoing improvement as technology evolves. At OpenAI, that commitment has shaped a number of recent initiatives focused on creating safer experiences for younger users. We recently introduced our Child Safety Blueprint, outlining policy proposals and practical steps to help improve protections for children online. This builds on the safeguards we already have in place: age prediction technology designed to help identify teen users so we can provide more protective, age-appropriate experiences with content protections to reduce exposure to sensitive or potentially harmful material. We offer parental controls so that parents can choose how AI shows up in their home. And through our teen safety policies for GPT-OSS-Safeguard, we’ve shared practical tools that can help others building AI systems implement stronger youth safety measures as well. There is always more work to do. We welcome rigorous external scrutiny and appreciate organizations like TeenAegis that are helping build measurement frameworks to encourage a race to the top on child safety. https://lnkd.in/gv6BBEvE

Introducing GPT-Rosalind, our frontier reasoning model built to support research across biology, drug discovery, and translational medicine. We believe advanced AI systems can help researchers move faster through the discovery process, not only by improving efficiency, but by helping scientists explore more possibilities, uncover connections that might otherwise be missed, and reach stronger hypotheses sooner. This model is optimized for scientific workflows, with stronger performance in protein and chemical reasoning, genomics analysis, biochemistry knowledge, and scientific tool use. It is designed to support evidence synthesis, hypothesis generation, experimental planning, and other multi-step research tasks that shape the earliest stages of discovery. Over time, systems like this could help scientific organizations uncover new breakthroughs faster and turn promising ideas into real-world advances sooner. Starting today, GPT-Rosalind, our Life Sciences model series, is available in research preview for qualified customers, including Amgen, Moderna, the Allen Institute, and Thermo Fisher Scientific, through ChatGPT, Codex, and the API. https://lnkd.in/g2TRfc5r