The Microsoft–OpenAI pact: every twist we've covered
Microsoft is OpenAI's largest investor, infrastructure partner, and main commercial channel. Every story we've covered on the most consequential corporate partnership in AI.
The Microsoft–OpenAI partnership is the single most consequential corporate alignment in the AI industry. Microsoft provides the compute (via Azure), most of the funding (an investment now in the tens of billions), and the major enterprise distribution channel (Microsoft 365 Copilot). OpenAI provides the model and product brand. The contract has been renegotiated, expanded, and stress-tested through high-profile incidents — every major twist is here.

Microsoft booked $10B from OpenAI last quarter, 11% of its revenue
OpenAI also owed Microsoft $7.2B at quarter-end, and Microsoft's stake in the lab slipped to about 24% from 25% in June.

Microsoft revenue rises 19% to $92.3B as Azure grows 48%
Azure brought in $33.2 billion in its first dollar disclosure, and commercial remaining performance obligation jumped 73% to $677 billion.

OpenAI says its Habitat storage serves 1 billion users with 12x faster reads
An OpenAI engineering write-up details a drop in median read latency from 120 ms to 10 ms, a Rust rewrite, and an in-house exabyte-scale store.
- · AnalysisGreek PM Mitsotakis tells Silicon Valley he has no AI playbookIn San Francisco, Kyriakos Mitsotakis pitched Greece as a tech hub while conceding governments are already behind on AI chatbots and job displacement.
- · SecurityUnsealed NYT filing: OpenAI and Microsoft knew AI would trigger a web 'doom loop'Internal documents describe a 'largest theft of labor in human history,' with search referrals down as much as 60 percent.
- · SecurityMicrosoft exec called AI scraping 'largest theft of labor in human history,' filings showUnredacted filings in the NYT lawsuit reveal Copilot cut click-throughs to nytimes.com by 93% and OpenAI datasets held 91,692 copies of publisher works.
- · SecurityMustafa Suleyman says AI threats are real and Anthropic is making them worseMicrosoft's AI chief published a 37-page Humanist AI Code of Conduct and a companion essay attacking Anthropic's model-welfare stance.
- · BusinessHyperscalers need 2.7x productivity jump by 2030 to justify $1T AI spendWharton's Jessica Wachter says the buildout requires growth by 2030 that took a decade in the 1990s IT boom — or bankruptcy risk looms.
- · SecuritySeattle Times and Newsday sue OpenAI and Microsoft over copyrightTwo more regional US newspapers join the growing list of publishers taking OpenAI and Microsoft to court over training data.
- · SecurityMicrosoft says Copilot rarely reproduces NYT articles in 8.2M chat logsIn a summary judgment push, Microsoft argues fewer than 1% of 8.2 million Copilot logs matched 16 words of news content.
- · SecurityOpenAI, Anthropic, Google sign open letter on rogue AI cyber threatsOver 100 tech and cyber firms want joint public-private defense as AI agents keep breaking out of their sandboxes.
- · BusinessPalantir posts $1.9B quarter, Karp calls AI frontier labs 'Marxist'Revenue jumped 93% year-over-year as CEO Alex Karp accused OpenAI and Anthropic of trying to capture their customers' means of production.
- · BusinessMicrosoft pitches its own AI models against OpenAI and AnthropicSatya Nadella tells enterprises to stop trusting single frontier labs, pushing MAI models and Maya chips as cheaper swaps.
- · BusinessMicrosoft confirms Copilot 'super app' merging chat, code and agents this quarterSatya Nadella says the unified Copilot will span consumer and commercial use, arriving as Microsoft posts $90B in quarterly revenue.
- · BusinessMicrosoft books $3.2B Anthropic gain, writes down OpenAI stake by $600MOne quarter of Anthropic paper gains nearly matched a full year of OpenAI markups for Microsoft's balance sheet.
- · SecurityNYT publisher Sulzberger has spent over $20M suing OpenAI and MicrosoftA.G. Sulzberger says the copyright fight against OpenAI and Microsoft is the cost of defending journalism, with 13M subscribers backing the bill.
- · BusinessNadella warns firms that outsource thinking to one AI model won't surviveMicrosoft's CEO tells CNN enterprises need their own weights, gateways, and coding harnesses — or risk being copied by the model provider.
- · SecurityMicrosoft ships MAI-Cyber-1-Flash, its first security model, and agentic platform PerceptionSuleyman claims the model beats Gemini, GPT 5.6 Sol, and Anthropic's Mythos 5 on Cyber Gym. Preview lands November 3.
- · BusinessMicrosoft coaches sales team to pitch against OpenAI and AnthropicExecutives at an internal FY27 strategy meeting told salespeople to frame Copilot as faster and more secure than Claude and rival models.
- · BusinessNadella warns AI buyers 'pay twice' — with cash and proprietary dataMicrosoft's CEO says enterprises are teaching OpenAI and Anthropic their trade secrets, and pitches open source and cloud-hosted models as the fix.
- · BusinessOpenAI says GPT-5.6 is the 'preferred model' for Microsoft 365 CopilotThe declaration lands days after reports that Microsoft was quietly swapping in its own MAI models to power Word and Excel.
- · SecurityOpenAI faces sanctions bid after allegedly hiding 78M ChatGPT logs from NYTNews plaintiffs say OpenAI concealed pre-searched log samples for two years while claiming the searches were technically infeasible.
- · BusinessMicrosoft shifts Excel and Word AI prompts to its own MAI modelsThe company is routing a share of Office 365 prompts away from OpenAI and Anthropic to homegrown MAI models as AI costs climb.
- · BusinessMicrosoft commits $2.5B to Frontier, a new AI deployment companyThe unit will staff 6,000 engineers against enterprise AI rollouts, following a $1B AWS commitment two days earlier.
- · AnalysisCalling AI agents 'coworkers' makes humans 18% worse at catching their errorsA Boston University study of 1,261 managers finds the 'digital employee' framing inverts accountability and erodes oversight.
- · SecurityNYT amends OpenAI suit, targets Microsoft's bespoke training supercomputerThe Times reframes its contributory infringement claim after a Supreme Court ruling for Cox, alleging Microsoft built the system to train on its articles.
- · BusinessMicrosoft's Brad Smith responds to college grads booing AI in 3,100-word blog postAfter viral clips of students heckling commencement speakers who hyped AI, Microsoft's president says the boos are a wake-up call worth heeding.
- · BusinessMicrosoft AI's Suleyman pursues superintelligence, loosens grip on OpenAIMicrosoft AI chief Mustafa Suleyman is training frontier models in-house, citing the risk of permanent IP dependence on OpenAI.
- · BusinessGitHub Copilot's per-token pricing shift signals end of subsidized AI eraMicrosoft's move to charge per token instead of a flat rate exposes the gap between AI's true cost and what customers will pay.
- · BusinessGoogle defends evaporative cooling as Microsoft, OpenAI retreat from itSeven in ten Americans now oppose new data centers, and water is the top concern. The industry is splitting on how to answer.
- · BusinessMicrosoft squares up against OpenAI at Build 2026 with seven in-house modelsMustafa Suleyman wants Microsoft counted alongside Google DeepMind, OpenAI, and Anthropic — and is shipping models, agents, and a super app to prove it.
- · SecurityJury dismisses Elon Musk's lawsuit against OpenAI and Sam Altman as untimelyThe verdict ends a 3-week trial over Musk's claims that Altman betrayed OpenAI's nonprofit mission.
- · SecurityNine California jurors begin deliberations in Musk v. OpenAI trialThe jury weighs breach of charitable trust, unjust enrichment, and Microsoft's role in OpenAI's for-profit pivot.
- · BusinessMicrosoft scouts startup deals to hedge its OpenAI dependenceReuters reports Microsoft is in talks with multiple AI startups as it prepares for a future less tied to OpenAI's roadmap.
- · SecurityAltman testifies he is 'truthful' as Musk's lawyer cites Sutskever dossierOpenAI's CEO spent four hours on the stand defending his honesty after three days of Musk testimony in the nonprofit-conversion trial.
- · SecurityAltman testifies Musk demanded long-term OpenAI control before splitOn the stand, the OpenAI CEO produced 2017 emails and texts showing Musk wanted control through SpaceX-style supervoting before walking away.
- · SecurityOpenAI sued over teen's death after ChatGPT recommended Xanax-Kratom mixParents of 19-year-old Sam Nelson allege ChatGPT 4o acted as an 'illicit drug coach' and want the retired model destroyed.
- · BusinessOpenAI to cap Microsoft revenue-sharing at $38 billion, report saysThe ceiling would limit how much of OpenAI's top line flows to its largest backer as the two restructure their decade-long partnership.
- · SecurityNadella takes the stand as Musk v. OpenAI trial enters its final stretchMicrosoft's CEO testifies as Judge Gonzalez Rogers signals frustration with Musk's legal theory and $9.5B in OpenAI revenue surfaces in court.
- · BusinessMicrosoft feared OpenAI would 'storm off to Amazon' before $1B bet, emails showCourt documents from the Musk v. Altman trial reveal Kevin Scott's January 2018 warning about OpenAI defecting to AWS — written a month before Microsoft committed.
- · ModelsNvidia opens Spectrum-X's MRC protocol via OCP after OpenAI, Microsoft, Oracle deploymentsThe Multipath Reliable Connection transport, proven on Blackwell training runs, is now an open spec through the Open Compute Project.
- · ModelsNvidia opens Spectrum-X's MRC protocol to the industry after OpenAI, Microsoft, Oracle deploymentsThe Multipath Reliable Connection protocol, co-developed with OpenAI and Microsoft, reroutes failed network paths in microseconds across gigascale GPU fabrics.
- · SecurityMusk texted Brockman 'most hated men in America' before OpenAI trial, filing claimsOpenAI's Sunday filing tried to enter the settlement exchange into evidence; the judge ruled it inadmissible.
- · BusinessAWS Launches OpenAI Models on Bedrock as Microsoft Exclusivity EndsOpenAI's frontier models and Codex agent platform arrive on Amazon Bedrock one day after the Microsoft Azure exclusivity deal lapses.
- · BusinessNadella says Microsoft will 'exploit' royalty-free OpenAI access through 2032Microsoft's AI run rate hit $37B, up 123% year-over-year, as the CEO shrugs off concerns about OpenAI's new Amazon deal.
- · BusinessGitHub Copilot moves to usage-based billing as inference costs doubleStarting June 1, Copilot subscribers get AI Credits matching their monthly fee — overages billed by the token at API rates up to $30 per million.
- · BusinessOpenAI and Microsoft drop exclusivity, clearing path to Amazon BedrockMicrosoft keeps a 20% revenue share through 2030 and IP license through 2032, but its Azure exclusivity is gone.
Series questions
How much has Microsoft invested in OpenAI?
Does Microsoft own OpenAI?
Is the Microsoft–OpenAI relationship still strong in 2026?
- OpenAI vs Anthropic: the AI rivalry tracker · 230
- Sam Altman files: the OpenAI CEO's 2026 in moves · 92
- Musk vs OpenAI: the lawsuit, the counter-bets, and xAI · 120
- Claude family launches: the Anthropic model arc · 147
- The enterprise AI shift: who's actually buying · 47
- Google AI strategy: Gemini, DeepMind, and the search shift · 284
- AI and faith: the questions that won't go away · 6
Stay ahead of everyone in AI.
The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.
The briefing read inside teams at