This project is suspended: no new articles or editions will be published. The archive stays available.
63 articles
OpenAI merged its Codex coding tool into a single ChatGPT desktop app and introduced ChatGPT Work, a new agent for long, multi-step office tasks, alongside the general release of GPT-5.6.
Read more →An internal audit found that roughly a third of SWE-Bench Pro's coding tasks are flawed, prompting OpenAI to withdraw its recommendation that developers rely on the benchmark.
Read more →Meta Superintelligence Labs opened a public preview of Muse Spark 1.1, a paid agentic coding model with a 1-million-token context window, taking on OpenAI and Anthropic.
Read more →OpenAI opened its GPT-5.6 model family — Sol, Terra and Luna — to all ChatGPT, API and Codex users on July 9, after weeks of government-mandated review limited access to select partners.
Read more →OpenAI has begun rolling out GPT-Live, new voice models that can listen and speak at once, replacing Advanced Voice Mode for all ChatGPT users starting July 8.
Read more →OpenRouter data shows Chinese AI models now handle over 30% of tokens run by US firms every week since February, peaking at 46%, as companies chase costs up to 90% lower than Anthropic or OpenAI.
Read more →OpenAI has released GPT-Realtime-2.1 and a smaller mini version for its Realtime API, cutting voice-agent response latency by at least 25% through improved caching.
Read more →Meta's superintelligence chief told employees that the company's next model, codenamed Watermelon, has caught up to OpenAI's GPT-5.5 on key benchmarks — though the specific tests and results have not been disclosed.
Read more →Microsoft committed $2.5 billion and 6,000 engineers to a new subsidiary that stations AI specialists directly inside client organizations, competing with similar offerings from OpenAI, Anthropic, and Amazon.
Read more →OpenAI has released GeneBench-Pro, a 129-problem benchmark that tests whether AI agents can exercise the judgment-intensive analytical reasoning required for real-world computational biology research.
Read more →OpenAI has proposed granting the U.S. government a 5% equity stake — worth roughly $42.6 billion at its current valuation — to ease regulatory pressure and share AI's economic gains with the public.
Read more →OpenAI launched GPT-5.6 Sol, Terra, and Luna on June 26, restricting initial access to roughly 20 government-approved organizations in the first application of a new US federal voluntary review framework for frontier AI models.
Read more →OpenAI is leaning toward pushing its stock market debut to 2027, with CEO Sam Altman rejecting any listing price below $1 trillion even as advisers flag tech market volatility.
Read more →Google has pushed its flagship Gemini 3.5 Pro model from June to July 2026, while a Nobel Prize winner and three other senior researchers departed for OpenAI and Anthropic in the same week.
Read more →OpenAI on June 22 expanded its Daybreak security program with GPT-5.5-Cyber, a model designed to detect and patch code vulnerabilities, and launched Patch the Planet to harden over 30 major open-source projects.
Read more →OpenAI unveiled three new models — Sol, Terra, and Luna — on June 26, but restricted access to government-vetted partners following a White House request on frontier AI oversight.
Read more →OpenAI retired GPT-4.5 from ChatGPT on June 26, removing the last GPT-4-era model from the consumer app after a 30-day notice period. API access remains unchanged.
Read more →OpenAI unveiled three new models — Sol, Terra, and Luna — under the GPT-5.6 banner, with access currently restricted to a small group of trusted partners cleared by the U.S. government.
Read more →Nobel laureate John Jumper and transformer pioneer Noam Shazeer both left Google DeepMind within 48 hours, heading to Anthropic and OpenAI as Alphabet's stock fell over 5%.
Read more →OpenAI and Broadcom revealed Jalapeño on June 24 — a purpose-built inference chip designed in nine months with AI assistance, targeting deployment by end of 2026.
Read more →OpenAI and Trail of Bits launched Patch the Planet, using the GPT-5.5-Cyber model to find and fix vulnerabilities in widely used open-source software, with 30+ projects on board.
Read more →