OpenAI Takes the Pentagon, Nvidia Rattles Markets, London Protests: March 2, 2026
OpenAI steps in to claim a controversial Pentagon defense contract as Anthropic is blacklisted, Nvidia's $30B OpenAI bet triggers a tech sell-off, and hundreds march through London's AI hub in the largest anti-AI protest yet.
OpenAI Seizes Pentagon Contract Hours After Anthropic Is Blacklisted for Refusing Weapons Guardrails
OpenAI swooped in to claim a lucrative Department of War contract on Saturday — reportedly worth around $200 million — just hours after the Trump administration blacklisted rival Anthropic for refusing to agree to deploy its AI without restrictions on mass surveillance and autonomous weapons. The Pentagon had demanded Anthropic remove contractual safeguards prohibiting those use cases; when Anthropic held its ground, the DoD terminated the contract and labeled the company a national security risk. OpenAI CEO Sam Altman quickly stepped in, later fielding a public Q&A on X to defend the decision — and was met with far more hostility than he expected.
The deal grants the military access to OpenAI models on a classified network, with Altman asserting the company's own red lines — no mass domestic surveillance, no directing of autonomous weapons systems — are more enforceable because deployment stays cloud-only with OpenAI personnel in the loop. Critics, including MIT Technology Review, pushed back, arguing OpenAI's "compromise" still permits the very activities Anthropic had refused to enable, just with softer framing.
The backlash spread quickly. ChatGPT users on Reddit and X began circulating guides to cancel subscriptions and migrate to Claude, and Anthropic's app shot to the top of the App Store in the days that followed. The episode lays bare a growing tension at the heart of the AI industry: as frontier AI labs become embedded in national security infrastructure, the ethical commitments that distinguish one company from another are being tested in real operational conditions — with no clear consensus on how far those lines should extend.
Nvidia's $30B OpenAI Gamble Sparks Market Sell-Off as Investors Demand AI Revenue Reality Check
Nvidia shares dropped roughly 4% on March 2, extending a slide that began after the chipmaker confirmed its $30 billion stake in OpenAI's record-shattering $110 billion funding round finalized in late February. The round — the largest private technology financing in history — values OpenAI at $730 billion, with Amazon contributing $50 billion and SoftBank another $30 billion. The deal cemented OpenAI's position at the center of the AI economy, but it has also unsettled institutional investors.
At the heart of the anxiety is what analysts are calling a "circular revenue" problem: OpenAI is contractually obligated to spend a significant portion of the new capital on Nvidia's next-generation Vera Rubin NVL72 GPU systems. Nvidia, in effect, invested in a company that will funnel that money back to Nvidia for hardware — a loop that raises uncomfortable questions about whether AI sector revenues are as organic as headline numbers imply. Despite Nvidia posting blowout quarterly earnings of $68.1 billion — a 73% year-over-year increase — the market's reaction reflected deeper unease.
The sell-off spread beyond Nvidia, weighing on the Nasdaq Composite and S&P 500 as investors signaled a broader shift in sentiment. The market appears to be moving from a "faith-based" phase of AI investing — where announcements alone drove valuations — into a "show me the revenue" era where sustainable unit economics matter. With Oracle having recently announced a $50 billion data center expansion and Microsoft continuing massive infrastructure commitments, the question is whether the AI industry can generate returns commensurate with the hundreds of billions being deployed.
Hundreds March Through London's AI Hub in the Largest Anti-AI Protest Yet
On February 28, several hundred demonstrators wound through London's King's Cross neighborhood — home to the UK offices of OpenAI, Meta, and Google DeepMind — in what organizers from Pause AI and Pull the Plug called the largest anti-AI protest ever held. Concerns ranged widely: AI-generated online slop, abusive deepfakes, mass unemployment, and fears of existential catastrophe all appeared on signs and sandwich boards. One participant wore a billboard on their head reading "WHO WILL BE WHOSE TOOL?" Another simply said: "Stop using AI."
MIT Technology Review journalist Will Douglas Heaven attended the march, describing an eclectic but growing crowd. The protest's central demand is clear — AI company CEOs should publicly commit to pausing training of the most powerful AI systems until adequate international safeguards are established. Organizers representing Pause AI, Pull the Plug, Mad Youth Organise, and Blaksox co-ordinated the event, emphasizing that the movement is no longer a fringe concern.
The growth trajectory of these protests is striking. In May 2023, only two or three demonstrators showed up outside a Sam Altman lecture in London; by last June, Pause AI drew a few dozen to a Google DeepMind office. Saturday's march of several hundred marks a significant escalation — and it arrives at a moment when AI's visibility in daily life, government contracts, and geopolitical strategy has never been higher. The movement's momentum suggests public pressure on AI governance is becoming a meaningful political force, even as governments remain divided on how to respond.
What You Can Do
Read OpenAI's Own Account of the Pentagon Deal
OpenAI published its rationale for the Department of War agreement, including its stated red lines on surveillance and autonomous weapons. Worth reading to form your own view on whether the safeguards hold.
Explore the Pause AI Movement
Pause AI is the international organization behind Saturday's London march. Their site explains the case for slowing AI development and how to get involved.
Track OpenAI's $110B Funding Round Fallout
Multiple outlets are covering the market implications of the historic OpenAI funding round and Nvidia's involvement. TechCrunch and Bloomberg have the most detailed analyses.