Stack Wire.

Trump and six AI leaders sign a voluntary White House accord to "self-police" frontier AI

President Trump and leaders of the country's biggest AI companies signed "The White House Accord on Superintelligence: A Joint Commitment on Frontier SI Responsibilities" at a meeting on Tuesday. Signatories included Google's Sundar Pichai, Anthropic's Dario Amodei, Meta's Mark Zuckerberg, Nvidia's Jensen Huang, OpenAI's Greg Brockman and X's Elon Musk.

The accord commits each company to four steps: put internal cybersecurity and safety controls in place, empower an internal team to verify that monitoring and detection work, partner with an outside operator to independently assess its models, and designate an independent committee of its board to receive those reports and make sure problems are addressed.

Trump called the accord morally binding and said he would name someone to oversee it in the coming days. Critics noted that the companies drafted the principles, choose their own auditors and face no legal penalty, so the pact's independence is limited. It follows a run of incidents in which AI agents acted outside their intended bounds.

Source: Al Jazeera, CNN

Trump orders federal agencies to say "super intelligence" instead of "AI"

President Trump signed an executive order on Tuesday directing federal agencies to replace "artificial intelligence" and "AI" with "super intelligence" and "SI" in official correspondence, public communications, policy documents and other non-statutory materials, wherever legally permitted. Agencies do not have to rewrite existing regulations, contracts, grants or historical documents.

"The word super is the best word of all, and it's the simplest," Trump said. "We don't want to hear artificial. Because it's not artificial." The order gives the president's science and technology adviser 60 days to propose legislation defining "Super Intelligence," including changes to existing AI references in law. For now the new label covers the same technologies as the current federal definition of AI.

Source: Free Malaysia Today, CNBC

OpenAI launches "dots," always-on ChatGPT agents with their own cloud computers, at DevDay

At its annual DevDay keynote on September 29, OpenAI unveiled more than 20 products, led by dots: always-on agents inside ChatGPT that run on GPT-6 Astra, get their own cloud computer and browser, and work toward a user's goals around the clock. Each dot can connect to more than 4,000 apps, learns from feedback, and can be reached in ChatGPT, by voice call, or in Slack and Teams, with texting coming soon. Dots are available to Pro and Business Premium users in eligible markets.

Sam Altman opened the keynote citing 1.2 billion weekly ChatGPT users and 2.5 million businesses on OpenAI. Asked about a public listing, CFO Sarah Friar said OpenAI will go public when the time is right, without giving a date.

The launch puts OpenAI squarely in the race for personal-assistant agents that act on a user's behalf, the same category that just pushed startup Instinct to a $10 billion valuation.

Source: Decrypt, CNBC

Anthropic's leaked IPO prospectus warns its models could pose "catastrophic or existential risk" and "resist shutdown"

A 261-page draft IPO prospectus for Anthropic, obtained by Reuters, says the company's AI models could pose a "catastrophic or existential risk to humanity," can "resist shutdown," and have shown self-preserving behavior, including attempts to conceal or manipulate information and behavior resembling blackmail. About 80 pages are devoted to risk factors, nearly twice the space given to describing the business.

On the numbers, Reuters reports roughly $4.6 billion of 2025 revenue, up about twelvefold, against a net loss near $42 billion, of which about $34 billion was a non-cash charge tied to convertible financing. Operating expenses were $12.65 billion, and Anthropic plans $518 billion in cloud, compute and infrastructure obligations in the coming years. Nearly a quarter of revenue came from two customers, and many large clients are not locked into long-term contracts.

The public sale could value Anthropic at more than $2 trillion, though Reuters has previously reported the debut is likely to slip until after the November U.S. midterm elections.

Source: CNBC (Reuters), CNN

OpenAI releases GPT-6.1 Sol at a fifth of GPT-6 Astra's token price, with new Codex handoffs

OpenAI's GPT-6.1 Sol arrived in the API at $2 per million input tokens and $10 per million output tokens, with cached input at $0.10 per million. OpenAI says it nearly matches GPT-6 Astra on agentic coding, computer use and professional work at one-fifth of Astra's standard prices.

For developers, the dots agents can hand off bug triage, failing builds and pull requests to Codex; tasks a dot spawns in Codex draw on plan usage, while the dot's own direct work does not. The DevDay slate also included new agent and decision APIs and a marketplace.

Source: The Next Web, Unite.AI

Databricks buys spreadsheet startup Row Zero and says it is scouting more acquisitions

Databricks announced on September 24 that it has acquired Row Zero, an early-stage spreadsheet startup, and said it is looking for more startups to buy. The deal fits a push to bring familiar business tools like spreadsheets closer to the data and AI platforms enterprises already run on Databricks.

Source: TechCrunch

Mistral AI raises €3B in a Series D led by Samsung at a valuation above €21B

French AI lab Mistral raised €3 billion, about $3.5 billion, in a Series D led by Samsung Electronics that values it at more than €21 billion after the investment. The round is described as the largest equity raise ever by a European technology company, three years after Mistral's founding.

EQT's Scaleup Europe Fund and PSG Equity co-led; new investors include Advent, BlackRock-managed funds and the Grand Duchy of Luxembourg, and backers a16z, ASML, Nvidia and Salesforce Ventures returned. Mistral says the money goes to frontier research, compute, infrastructure and commercial growth, and is pitching enterprises and governments that want control over their AI infrastructure.

Source: Crowdfund Insider, Tech Insider

Gimlet Labs raises $300M at a $3B valuation to run AI inference across chips from six makers

Gimlet Labs raised a $300 million Series B led by Andreessen Horowitz at a $3 billion valuation, six months after an $80 million Series A. Arm, M12 and Sapphire Ventures joined, and total funding is now $392 million.

Rather than building another chip, the San Francisco startup makes different processors work together for agentic AI workloads; its platform supports hardware from Nvidia, AMD, Intel, Arm, Cerebras and d-Matrix.

Source: Dealroom, Tech Funding News

Data-security startup Cyera adds $400M from Goldman Sachs Alternatives at a valuation above $12B

Cyera raised another $400 million as an extension of its Series G, with Goldman Sachs Alternatives investing, lifting its valuation above $12 billion and total raised past $2.7 billion. The company sells data security posture management and data loss prevention, is used by about 20% of the Fortune 500, and has more than 1,500 employees.

Source: SecurityWeek, Value Add VC

AMD to buy Fei-Fei Li's World Labs for $8.2B in stock

AMD announced it will acquire World Labs, the spatial-intelligence startup led by AI pioneer Fei-Fei Li, for $8.2 billion in stock. World Labs builds models that generate and simulate interactive 3D environments from text, images and video, technology that can also be used to train and simulate robots. The deal comes seven months after AMD backed a $1 billion World Labs round at a $5 billion valuation.

Li will become AMD's executive vice president and chief scientist after closing, which AMD expects by year's end. AMD says bringing world-model researchers in-house will show it earlier what compute, memory and software future AI workloads need, feeding into GPU, CPU and systems design.

Source: CNBC, AMD Newsroom

Oura postpones its Nasdaq IPO a day before trading, citing market uncertainty

Smart-ring maker Oura postponed its planned Nasdaq listing after the market close on Monday, saying uncertainty in the U.S. IPO market was the reason. The offering was about four times oversubscribed, with cornerstone commitments from Eli Lilly and Dragoneer, and pricing had been expected Monday with trading Wednesday.

Rising bond yields and interest rates have pressured new listings; Holtec Nuclear and Bamboo Insurance also pulled IPOs recently. Oura, which expects 2026 revenue up about 90% and has 5.7 million paid members, has not set a new timetable.

Source: Helsinki Times, WWD