OpenAI

On July 9, 2026, OpenAI moved GPT-5.6 Sol, Terra, and Luna from a government-coordinated trusted-partner preview (started June 26) to broader public availability after U.S. officials cleared a wider rollout. The three-tier family—Sol flagship, Terra balanced (~2× cheaper vs GPT-5.5-class positioning), and Luna low-cost—had been gated at the administration’s request.

On July 8, 2026, OpenAI published an audit finding roughly 30% of SWE-Bench Pro tasks broken—hidden requirements, contradictory instructions, overly strict tests, or incomplete grading—and retracted its prior recommendation that the research community treat SWE-Bench Pro as a leading coding evaluation.

On July 8, 2026, OpenAI introduced GPT-Live—a new generation of full-duplex voice models powering ChatGPT Voice. GPT-Live-1 (default for Go/Plus/Pro) and GPT-Live-1 mini (default for Free) can listen and speak simultaneously, with global consumer rollout on web and mobile; Business/Enterprise/Edu excluded at launch.

On July 2, 2026, the Financial Times and major wires reported that OpenAI is in early talks to give the U.S. government a 5% stake—worth roughly $40B+ at recent private valuations—as part of a broader idea that leading U.S. AI labs would allot equity to a public vehicle modeled on funds like Alaska’s permanent fund. CEO Sam Altman framed the concept as sharing AI’s benefits with Americans amid Washington pressure.

On June 26, 2026, OpenAI began a limited preview of the GPT-5.6 series—Sol (flagship), Terra (balanced, ~2× cheaper than GPT-5.5-competitive performance), and Luna (fast, lowest cost)—initially via API and Codex for a small set of trusted partners whose participation was shared with the U.S. government. OpenAI said broader availability is planned in coming weeks and that it does not want government access gating to become the long-term default.

On June 25, 2026, The New York Times reported that OpenAI is leaning toward holding its initial public offering until 2027 rather than as early as late 2026, according to three people involved in the company’s deliberations. Advisers presented options of waiting for a ~$1 trillion valuation in 2027 or accepting a lower valuation for a faster listing—with Sam Altman described as treating a cut to the trillion-dollar target as a nonstarter.

On June 24, 2026, OpenAI and Broadcom unveiled Jalapeño—OpenAI’s first “Intelligence Processor,” a custom accelerator designed from a blank slate for LLM inference. The chip is the first in a multi-generation compute platform co-developed with Broadcom (silicon/networking) and partners such as Celestica for boards/racks, with initial deployment targeted by the end of 2026 and early testing claimed to beat current state-of-the-art performance per watt.

On June 22, 2026, OpenAI released updates under its Daybreak cybersecurity program: the full version of GPT-5.5-Cyber for trusted defenders (85.6% CyberGym vs 81.8% for GPT-5.5) and Patch the Planet, an initiative with partners including Trail of Bits to help open-source maintainers find, validate, and fix vulnerabilities with AI. Early results cited hundreds of issues and dozens of merged patches across major projects.

On June 21, 2026, OpenAI announced that Samsung Electronics is deploying ChatGPT Enterprise and Codex to employees worldwide—one of OpenAI’s largest enterprise rollouts. Coverage will include all Samsung Electronics employees in Korea and all Device eXperience (DX) division employees globally, reversing Samsung’s 2023-era caution after earlier generative-AI data-leak concerns.

On June 18, 2026, Dean W. Ball announced he will join OpenAI on July 6 as Head of Strategic Futures—a new team reporting to Chief Strategy Officer Jason Kwon that will shape frontier AI policy on catastrophic risk, recursive self-improvement, labor markets, and lab–government relations. Ball, a co-author of the Trump administration’s America’s AI Action Plan and Nonresident Senior Fellow at the Foundation for American Innovation, said Hyperdimensional writing remains independent without OpenAI preapproval.

On June 17, 2026, OpenAI began rolling out a new Scheduled tasks experience in ChatGPT, with a dedicated sidebar Scheduled page to view, pause, resume, edit, and delete automated jobs. Users can ask ChatGPT to send reminders, handle recurring work, or monitor for changes, with scheduling for specific times or broader windows (morning/afternoon/evening). The update targets Plus, Pro, Business, and Enterprise users, with tier-based active-task limits.

On June 17, 2026, Noam Shazeer—Google vice president of engineering and co-lead of Gemini, Transformer co-author, and Character.AI co-founder—announced he is leaving Google to join OpenAI. The move comes less than two years after Google’s reported $2.7 billion Character.AI-linked rehire that returned him to lead Gemini work, and as OpenAI advances its confidential IPO process.

On June 14, 2026, OpenAI introduced the OpenAI Partner Network, its first formal global partner program, backed by $150 million in investment. The program targets consulting firms, systems integrators, and technology companies that build, sell, and deliver OpenAI solutions—with a stated goal of certifying 300,000 AI consultants by the end of 2026—deepening OpenAI’s enterprise go-to-market alongside its confidential S-1 process.

On June 10, 2026, OpenAI published research on PRC-linked influence operations that used ChatGPT to generate content aimed at shaping U.S. public and political debate over AI infrastructure, data centers, and related trade issues such as tariffs. The company said it detected the campaign because operatives relied on its models to produce social-media-ready material, adding a new dimension to how AI tools themselves become instruments in information warfare around AI policy.

On June 10, 2026, OpenAI announced that enterprise customers can access OpenAI models and Codex using existing Oracle Universal Credits (UCM) on Oracle Cloud Infrastructure. The partnership lets teams draw down Oracle cloud commitments for frontier model usage, agentic coding, and AI application workloads without a separate OpenAI-only procurement path—expanding OpenAI’s multi-cloud distribution beyond Azure-centric narratives.

On June 8, 2026, OpenAI announced it had confidentially submitted a draft Form S-1 registration statement to the U.S. Securities and Exchange Commission. In a short company post, OpenAI said it expected the filing to leak so it was announcing it; timing of any IPO remains undecided because some goals may be easier as a private company, but the filing creates the option to go public sooner if that proves best. The move comes seven days after Anthropic’s confidential S-1 and amid SpaceX’s mega-IPO process.

In early June 2026, OpenAI rolled Lockdown Mode more broadly across personal ChatGPT accounts and self-serve Business workspaces, with major press coverage peaking June 6. The optional security setting deterministically disables or limits tools that could be abused for data exfiltration via prompt injection—including outbound network behaviors tied to browsing and some agent capabilities—while OpenAI cautions that residual injection risk remains in cached content and uploads.

On June 4, 2026, OpenAI began rolling out Dreaming V3, a major ChatGPT memory architecture that synthesizes user context in the background without requiring users to explicitly save memories. The system first reached ChatGPT Plus and Pro users in the United States, with expansion to more countries and Free/Go tiers planned over subsequent weeks. OpenAI positions dreaming as roughly 5× more compute-efficient than prior designs, enabling broader availability.

On June 3, 2026, OpenAI announced new capabilities for GPT-Rosalind, its enterprise life-sciences model series. The update pairs GPT-5.5–class agentic coding and tool use with improved performance on drug discovery, medicinal chemistry, genomics, analysis, design, and experimental workflows, while secondary reporting cites roughly 31% fewer tokens than GPT-5.5 on comparable workloads.

On May 29, 2026, OpenAI announced Rosalind Biodefense, expanding trusted access to GPT-Rosalind—its frontier life-sciences reasoning model—for vetted developers and U.S. government partners working on biodefense, public health, and pandemic preparedness. The program is framed as defensive acceleration in biology, pairing capability access with trust boundaries rather than fully open deployment of dual-use life-sciences AI.

On May 29, 2026, OpenAI’s ChatGPT release notes detailed major Codex updates: Computer Use support on Windows in the Codex app so eligible users can have Codex see, click, and type in Windows applications; remote control so work started on a Windows host can be monitored and steered from ChatGPT on iOS/Android or Codex on Mac; infrastructure improvements for responsiveness and in-app browser stability; and Codex Profiles showing identity, activity, usage stats, and token activity. Computer Use on Windows is unavailable in the EEA, UK, and Switzerland at launch.

On May 27, 2026, OpenAI published its election-year safeguards plan for 2026, the second major global election cycle of the generative AI era. The program includes five planks: reliable voting and results information (AP live vote counts for the U.S. and Brazil; Democracy Works for U.S. registration and polling logistics), cybersecurity support for election defenders, deepfake watermarking and provenance (SynthID and C2PA), enforcement against election interference and scaled campaign messaging, and efforts to keep ChatGPT politically neutral by default.

On May 19, 2026, OpenAI announced Guaranteed Capacity, a new program allowing eligible customers to secure long-term access to OpenAI compute for AI products, agents, and workflows through 1-3 year commitments with increasing discounts based on annual spend levels.

On May 18, 2026, at Dell Technologies World, OpenAI and Dell Technologies announced a partnership to deploy Codex — OpenAI's autonomous coding agent — inside hybrid and on-premises enterprise environments via the Dell AI Data Platform and Dell AI Factory, enabling secure use where enterprises' data, systems, and workflows already live.