A Spy at the Center of the Gulf's AI Chip Deal
Today's most striking story reaches into the shadows of intelligence work. A Wall Street Journal report, citing sources, says a CIA operative played a dual role in one of the Gulf's biggest technology wins — the United Arab Emirates' path to securing American AI chips. The Journal identifies the operative as Jonny Gannon, who it reports was involved in vetting the UAE's access. The account underscores how deeply the machinery of national security has become entangled with the global race for advanced semiconductors, the scarce hardware that powers modern artificial intelligence. When a spy is helping clear the way for a foreign partner to buy American chips, the line between commercial technology and statecraft has all but disappeared.
SpaceX Pushes Into Pentagon Computing
Elon Musk's SpaceX, best known for rockets and satellites, is reportedly in talks to sell computing power to the U.S. military — a move that would carry the company deep into the national-security side of artificial intelligence. According to a Wall Street Journal report cited by Reuters, the discussions center on a multi-billion-dollar agreement to supply the Department of Defense, drawing on Musk's Colossus AI supercomputer.
It's a notable expansion of ambition. SpaceX already sits at the intersection of space and defense through its launch and satellite work; supplying raw AI compute to the Pentagon would add a new and lucrative dimension. For the military, access to a system like Colossus could accelerate everything from data analysis to model training. For SpaceX, it deepens a relationship with the government that has already made it a cornerstone of America's space program — and now, potentially, of its AI infrastructure too.
The Nuclear Warning
Against that backdrop of accelerating military AI, a sharp note of caution arrived today. A prominent anti-nuclear campaigner, in an interview published by the international press agency Pressenza, warned that mixing artificial intelligence with the world's nuclear arsenals could be a recipe for catastrophe. The message — "Don't let AI get the bomb" — is a plea to keep human judgment firmly in control of the most consequential decisions humanity has ever built. As AI systems grow more capable and more embedded in defense planning, the campaigner's warning is a reminder that speed and automation carry real dangers when the stakes are nuclear.
Pushback on the Ground
The unease isn't confined to interviews and op-eds. In the Pacific Northwest, defense-technology company Anduril — known for its autonomous systems and AI-driven weapons — is facing organized opposition. According to GeekWire, activists are planning a weekend protest at Anduril's Seattle office, taking aim at the company's role in developing AI weapons. It's a visible sign that public debate over autonomous and AI-enabled warfare is spilling into the streets, even as the industry's biggest players race to win military contracts.
The Bigger Picture
Taken together, today's developments sketch a single arc. Governments and companies are moving fast to fuse artificial intelligence with national defense — securing the chips that make it possible, courting the compute that makes it powerful, and building the autonomous systems that put it to use. At the same time, campaigners and activists are pushing back, questioning where the guardrails are and who is watching them. The tension between those two forces — acceleration and restraint — is shaping up to be one of the defining stories of the year in defense technology.