Topic
Geopolitics
On the US-China technology competition, the frontier lab leaders are broadly aligned. Anthropic's formal submission on the Commerce Department's diffusion rule (January 2025) argued that chip export controls are working, citing DeepSeek's own admission that restrictions are their primary constraint and noting that China must spend 2-4x more power to achieve comparable results. The submission pointed to a stark number: the U.S. share of global semiconductor production has dropped from 40% in 1990 to 12% today, making controls on advanced AI chips essential to prevent frontier training infrastructure from migrating overseas. Demis Hassabis has acknowledged the competition with China directly, though he frames it less as a zero-sum arms race and more as a question of sustained research quality and responsible deployment. Sam Altman, meanwhile, has been courting India aggressively, meeting with Prime Minister Modi in February 2026 and urging Indian audiences to build across the entire AI stack (from chips to applications), calling India OpenAI's fastest-growing market for its Codex coding tool.
The domestic front is where consensus breaks down. In late February 2026, Dario Amodei refused the Pentagon's demand that Anthropic drop its restrictions on two use cases: AI-driven domestic mass surveillance and fully autonomous weapons. The Department of War gave Anthropic a three-day ultimatum, threatening to designate the company a "supply chain risk" and invoke the Defense Production Act (tools normally reserved for foreign adversaries). Amodei held firm, arguing that AI-powered bulk analysis of Americans' personal data is "getting ahead of the law" and that current AI systems are "nowhere near reliable enough" for weapons that fire without human involvement. This was a striking position for a company that had been first to deploy models on classified government networks and first to build custom models for national security. Jack Clark, also at Anthropic, took a different angle on government adoption: he predicted AI would penetrate national security agencies quickly but warned of political paralysis in civilian government, forecasting a movement to freeze jobs in "bureaucratic amber" driven by "the chaotic winds of political forces." On the other side, Greg Brockman made headlines for significant political donations aligned with MAGA, suggesting OpenAI's leadership sees a different path to government relations. Altman has stayed more diplomatic, telling audiences that the main topics world leaders raise with him are infrastructure, jobs, and safety (in that order). Hassabis has pushed international cooperation through DeepMind's National Partnerships for AI program (launched with the US, UK, and India), but conceded that only governmental action could compel the whole industry to coordinate on safety. The overall picture: export controls and the China question produce agreement, but the relationship between AI companies and their own government is fracturing along lines that track with each company's theory of political power.
People on this topic
Perspectives
Amodei vs. Altman: The Pentagon Deal
When the Pentagon demanded unrestricted access to frontier AI, Dario Amodei refused and got blacklisted. Sam Altman said he agreed with Anthropic's red lines, then struck his own deal with the Department of War that same Friday night. The substantive disagreement is narrow but real: Amodei argued that existing law hasn't caught up with AI's ability to aggregate public data into comprehensive surveillance profiles, so the Pentagon's assurance that it would follow current statutes wasn't enough. Altman accepted that assurance, framing the deal as the Pentagon agreeing to OpenAI's principles. Seventy OpenAI employees signed a letter supporting Anthropic before Altman's deal went through. The episode crystallized the difference between the two leaders. Amodei treats safety commitments as constraints that must hold even when they're expensive, though his own company dropped its Responsible Scaling Policy pledge that same month under competitive pressure. Altman treats them as negotiating positions, things you advocate for but ultimately resolve through dealmaking rather than confrontation. Both approaches have costs. Amodei lost a major government contract and faces a supply-chain-risk designation. Altman kept the contract but earned the accusation that OpenAI replaced a blacklisted competitor while claiming solidarity with it.