OpenAI introduced GPT-Rosalind on April 16, 2026, a frontier reasoning model aimed at drug discovery, genomics analysis, protein reasoning, and scientific research workflows. The release extends the company’s model lineup into a more specialized scientific use case.
OpenAI on April 14 expanded its Trusted Access for Cyber program and introduced GPT-5.4-Cyber for vetted defenders, a move that broadens access to more permissive security tooling while keeping the rollout limited to verified users and teams.
OpenAI updated ChatGPT on March 24 with more visual shopping results, image-based product discovery and side-by-side comparisons, sharpening the app’s role as a commerce and research tool.
OpenAI has updated ChatGPT with GPT-5.3 Instant Mini as the fallback model after users hit rate limits for GPT-5.3 Instant, while also introducing a new $100-a-month Pro plan that expands Codex usage for heavier users. The changes, announced in the April 9 release notes, show how OpenAI is packaging ChatGPT more tightly around both conversational use and developer workflows.
OpenAI said on April 9, 2026, that CyberAgent is expanding its use of ChatGPT Enterprise and Codex, giving the chatbot a more formal place inside engineering and productivity workflows at one of Japan’s prominent internet companies.
OpenAI acquired TBPN on April 2, 2026, adding a live tech talk show to its growing mix of products, partnerships and distribution channels. The deal gives the company a louder voice in the ecosystem that tracks its own business.
Google has launched Veo 3.1 Lite, its most cost-effective video generation model, into the Gemini API and Google AI Studio, while also saying it will lower pricing for Veo 3.1 Fast on April 7. The move is aimed at developers building higher-volume video applications and marks a clear bid to widen adoption by reducing cost pressure in one of AI’s most compute-intensive formats.
Anthropic has disclosed that its annualized revenue run rate has passed $30 billion and that it will secure roughly 3.5 gigawatts of Google TPU capacity through a Broadcom-linked supply arrangement starting in 2027, a sign that frontier AI is now being measured as much by infrastructure scale as by model quality.
OpenAI said a third-party developer-tool compromise involving Axios forced it to update macOS app certification and require users to move to the latest desktop builds, while stating it found no evidence of user data access or software tampering.
OpenAI has introduced ChatGPT Health, a new ChatGPT experience built around health and wellness questions, marking a notable step deeper into one of the chatbot’s most sensitive and heavily used categories.
