AI News from OpenAI Blog

Latest coverage from OpenAI Blog, summarized and scored for signal.

• Perplexity trusts GPT-6 Astra with end-to-end systems — Perplexity is running OpenAI's GPT-6 Astra autonomously across communications, software changes, and production system monitoring — with far less human oversight than earlier models, marking a concrete step toward agentic AI operating at infrastructure level.
• Cognition helps Devin test its own work with GPT‑6 Astra — Cognition's Devin AI coding agent now uses GPT-6 Astra to test its own code changes, generating visual proof of working features and reducing the code engineers must review before shipping.
• How a researcher uses Codex and ChatGPT to search for new antimicrobial molecules — César de la Fuente's lab is using OpenAI's Codex and ChatGPT to mine living and extinct genomes for antimicrobial peptides, targeting drug-resistant infections at a time when antimicrobial resistance kills an estimated 700,000 people annually and threatens to become a leading cause of death this century.
• Now everyone can put data to work — OpenAI launched a Data agent inside ChatGPT Work that lets enterprise employees connect company data sources and build interactive dashboards using plain language — no SQL or dedicated BI tools required.
• Introducing ChatGPT for Financial Services — OpenAI launches ChatGPT for Financial Services, a purpose-built product combining proprietary financial data with GPT-6 Astra for investment research, financial modeling, and client-ready deliverables — OpenAI's first targeted enterprise vertical beyond general coding and chat.
• Expanding AI access and cyber defense for federal, state, local, and tribal governments — OpenAI and the U.S. General Services Administration will offer eligible federal, state, local, and tribal governments zero-cost licenses and 50% off usage fees, plus expanded cyber defense support — the most significant disclosed government AI procurement agreement from a frontier lab to date.
• Build more natural voice experiences with GPT‑Live‑1 in the API — OpenAI releases GPT-Live-1 in the API — a full-duplex voice model with stronger instruction following, custom voice support, and telephony integration, enabling developers to build low-latency conversational AI products directly connected to phone systems.
• Introducing the Agents API — OpenAI launched the Agents API, a managed cloud service built on the Codex harness that lets developers create and deploy long-running AI agents with tool use and orchestration — entering direct competition with AWS Bedrock Agents and Google Vertex AI Agent Builder.
• Paul Christiano joins OpenAI Foundation Board — Paul Christiano — founder of the Alignment Research Center and one of the most influential AI alignment researchers — joins the OpenAI Foundation Board and its Safety and Security Committee, adding independent safety expertise to OpenAI's governance structure.
• The AI policy window is open. We need to act. — OpenAI's VP of Global Affairs Chris Lehane publicly calls for binding AI policy frameworks, arguing the current legislative window won't stay open indefinitely and that expanding AI capabilities must be matched with stronger safety evidence and cross-company standards.
• GPT-6 Astra: The next generation in intelligence for work — OpenAI releases GPT-6 Astra, its most capable model for business use, with advanced reasoning, computer use capabilities, and stronger writing and design judgment — the first GPT-6 model explicitly scoped for enterprise and professional workflows.
• How GPT-5.6 Sol helps run quantum computing experiments — OpenAI showcases GPT-5.6 Sol paired with Codex autonomously running quantum computing experiments at MIT — calibrating qubits, analyzing results, and iterating without constant researcher intervention.
• Introducing ChatGPT Images 2.5 — OpenAI launched ChatGPT Images 2.5, an upgrade to its in-app image generation that improves personalization when working from text prompts, sketches, and reference photos — though the announcement provides no benchmark data or technical specifics.
• On the Navier–Stokes Millennium Prize Problem — OpenAI has published an AI-generated solution to the Navier–Stokes Millennium Prize Problem — one of seven Clay Mathematics Institute problems carrying a $1M prize — including a formal proof verified in Lean, potentially marking the first AI contribution to a solved Millennium Prize-level mathematical challenge.
• Funding grants for new research into AI and teen development — OpenAI launched a $5 million grant program funding independent research on how generative AI affects teen development, well-being, and safety—a strategic move to build an external evidence base on youth AI impacts ahead of anticipated regulation.
• 1Password increases engineering productivity 21% with Codex — 1Password reports a 21% increase in engineering productivity after adopting OpenAI's Codex for building new features and internal tools — a notable adoption signal from a security-critical company where code quality standards are non-negotiable.
• Supporting independent journalism in Ukraine — OpenAI teams up with AIRPPU and WAN-IFRA to launch an AI program for Ukrainian news organizations, targeting resilience and innovation capacity for independent journalism under wartime conditions.
• An Alien Mind — OpenAI Chief Scientist Jakub Pachocki publishes a reflective essay framing advanced AI as an 'alien mind' — cognitively foreign rather than a sophisticated tool — and calls for stronger alignment safeguards and international coordination as models grow more capable.
• Research acceleration: The view inside OpenAI — OpenAI shares early internal data on how coding agents are accelerating its own AI research — metrics on agent usage, experiment velocity, and task complexity offer a rare inside look at how the lab applies its own tools to scientific work.
• Daybreak for Frontline Defenders: $1B to protect essential services — OpenAI commits $1 billion through its Daybreak initiative to expand access to frontier cyber AI tools, training, and support for organizations defending critical infrastructure — hospitals, utilities, and government services.
• Legora reviewed 41 documents in minutes with GPT-6 Astra — OpenAI case study: legal tech platform Legora used GPT-6 Astra to review 41 financial documents in minutes, accurately identifying all four deliberately planted errors and improving overall workflow performance by nearly 40%.
• Playco cut manual fixes 50% prototyping games with GPT-6 Astra — OpenAI case study: game studio Playco used GPT-6 Astra to build three themed game prototypes from a single grey-box foundation, reporting 50% fewer manual fixes compared to the prior model generation.
• GPT-6 Astra: A new generation of intelligence — OpenAI's GPT-6 Astra debuts as the company's most capable and aligned model yet, with headline strengths in computer use, coding, cybersecurity, and science — a landmark flagship release that is light on technical specifics but reshapes AI competitive dynamics immediately.
• Safety overview: GPT-6 Astra — GPT-6 Astra — OpenAI's most capable broadly deployed model — becomes the first AI system to reach the 'Critical' cybersecurity capability threshold under OpenAI's Preparedness Framework, triggering mandatory safety mitigations before public rollout.
• ATV Big Air Tour turned 3 days of work into 3 hours with ChatGPT — OpenAI's ChatGPT compressed ATV Big Air Tour's 3-day marketing and merchandising workflow into 3 hours, including building an inventory website from merchandise photos in 15 minutes — a concrete SMB productivity case study.