The Omniscient Bulletin · October 9, 2026
OpenAI's revenue comes in lower than reported as Washington courts and penalizes Big Tech
OpenAI told investors its annualized revenue was about $50 billion, well below the figure widely reported last month, and AI chip and cloud stocks fell on Thursday. Washington courted and punished the industry on the same Thursday: eleven companies pledged $2.4 billion in compute credits and science tools to the Genesis Mission while the Labor Department suspended Microsoft and Adobe from its green card labor certification program. And Yandex cannot yet say whether a drone-struck data center housing two of its AI supercomputers can be restored.
Industry
OpenAI told investors its annualized revenue was about $50 billion at the end of September, below the $68 billion figure widely reported last month
OpenAI told investors its annualized revenue was roughly $50 billion at the end of September, CNBC confirmed, below the $68 billion that circulated widely in late September. That larger number counted gross revenue earned by OpenAI's partners, which makes the company easier to compare with Anthropic, according to a person familiar with the matter. The Financial Times had the $50 billion figure first. On Thursday, Nvidia shares fell 3%, Oracle nearly 6% and CoreWeave nearly 8%.
Labor
The Labor Department suspended Microsoft, Adobe and six IT outsourcing firms from its green card labor certification program
Labor Secretary Keith Sonderling said Thursday the department is barring Microsoft, Adobe, Cognizant, Infosys, Capgemini, Tata, Wipro and HCL from its Permanent Labor Certification program, the green card labor program. He said Microsoft and Adobe were suspended due to multiple active federal investigations, and that since 2009 the eight firms have received over 230,000 H-1B approvals. Microsoft says 80% of its 6,000 or so H-1B applications last fiscal year went to extensions or status changes for people it already employs.
Policy
Eleven companies pledged $2.4 billion in compute credits and science tools to the federal Genesis Mission, led by $1 billion from Nvidia
The White House said Thursday that eleven companies committed $2.4 billion in science tools and compute credits to the Genesis Mission Consortium, backing over 15 federal agencies working on national science and technology challenges. Nvidia committed $1 billion, AMD $500 million, OpenAI $200 million, and Anthropic and Google $150 million each, with $100 million each from AMP and Emerald AI and $50 million each from AWS, Armada, Crusoe and Micron.
Frontier
Anthropic launched OSS Scanner, a free opt-in service that sends open-source projects vulnerability reports generated entirely by its models, without human review
Projects that join Anthropic's OSS Scanner get periodic security scans by its strongest models at no cost, and Anthropic says the reports will be "fully model-generated, without human review or triage," which it says means some may be incorrect. Anthropic says it has found over 29,000 candidate vulnerabilities in six months but could manually triage only about 6,000. In validation, testers checked 97 critical and high-severity findings across 48 projects, and 85 (88%) met the bar for its disclosure process.
Frontier
The three safety researchers OpenAI fired last week wrote to its safety committees that they did not leak to The Information, and that the firings chill its culture
Tomek Korbak, Jasmine Wang and Mikita Balesni, the safety researchers OpenAI fired last week, wrote to OpenAI's Safety and Security Committee, Safety Advisory Group and Mission Advisory Council that they did not leak to The Information about less monitorable architectures. They urged OpenAI to keep last month's commitment to embed third-party safety auditors, and wrote that firings executed so abruptly are chilling its open culture. An OpenAI spokesperson told TechCrunch the firings followed a pattern of misconduct.
Policy
Anthropic dropped its blanket ban on personalized vote and campaign targeting, and added a prohibition on sustained abuse of its models, effective November 12
Anthropic's updated Usage Policy removes its "blanket prohibition on personalized vote and campaign targeting," which it says swept in legitimate civic work such as election officials sending ballot cure notices, and renames its elections section Do Not Undermine Democratic Processes. Deceptive political and commercial activity now sits in one new section on deceptive campaigns. The update also clarifies its weapons rules and bans sustained, needless abuse of its models in extreme cases. It takes effect November 12.
Compute
Yandex says it cannot yet confirm whether equipment at its drone-struck Sasovo data center, home to two of its three AI supercomputers, can be restored
Yandex said Thursday that the scale of the damage at its Sasovo data center in Russia was still being assessed and that it could not yet confirm whether the equipment there can be restored, after a Ukrainian drone strike started a fire that suspended operations at the site. Sasovo hosts two of the three supercomputers Yandex uses to build its AI model; Yandex hasn't said whether they were damaged. Zelenskiy did not mention Yandex but said Ukraine was mirroring Russian attacks on data centers.