Geopolitics

You know an AI model is serious when it gets two superpowers to pick up the phone. Anthropic’s Claude Mythos — the model so good at finding software vulnerabilities that it treats global infrastructure like tissue paper — just did something no diplomat, treaty, or summit could manage. It’s pushing the US and China toward actual AI safety talks. Not the performative kind. The “we’re both terrified” kind. The Model That Broke the Silence Quick recap if you’ve been living under a rock: Mythos can autonomously discover and exploit software vulnerabilities at a scale that makes seasoned security researchers nervous. The UK’s AI Safety Institute confirmed it can “execute multi-stage attacks on vulnerable networks and discover and exploit vulnerabilities autonomously.” Anthropic says it’s found holes in every major browser and operating system. ...

You can’t embargo the most valuable commodity on Earth and expect everyone to play nice. US prosecutors just dropped their biggest AI hardware enforcement action yet — a $2.5 billion smuggling operation that funneled restricted Nvidia servers through Bangkok directly into Chinese hands. The details read like a spy thriller. The implications hit like a freight train. The Operation Bangkok-based OBON Corp allegedly purchased massive quantities of Super Micro servers packed with Nvidia’s H200 and B300 chips — the silicon that powers frontier AI training. Instead of deploying them in Thailand’s growing AI ecosystem, the servers were rerouted to China. Alibaba is named as an end customer. ...

Beijing just dropped a one-line bomb on the global AI industry. China’s National Development and Reform Commission ordered Meta to unwind its $2 billion acquisition of Manus — the AI agent startup that was supposed to be Zuckerberg’s secret weapon. No negotiation. No diplomatic hedging. Just: reverse the deal. This isn’t a regulatory hiccup. It’s the moment the AI race officially split into two separate universes. What Made Manus Worth $2 Billion Manus builds general-purpose AI agents — software that doesn’t just chat but acts. It codes applications, runs market research, manages data analysis, and prepares budgets autonomously. Think of it as the generation after chatbots. ...

The biggest AI summit of 2026 just wrapped in New Delhi, and the headline isn’t a new model or a benchmark war. It’s this: 88 countries — including the US, China, and Russia — just signed a shared vision for how AI should be built and governed. That’s not supposed to happen. The Declaration Nobody Expected The India AI Impact Summit 2026 ran February 16–21 at New Delhi’s Bharat Mandapam. Prime Minister Modi inaugurated it alongside Macron and UN Secretary-General Guterres. The guest list read like a tech industry yearbook: Pichai, Altman, Amodei, plus delegations from over 100 countries and 20-plus heads of state. ...

Five days in New Delhi. $260 billion in pledges. A new US-India tech alliance with China squarely in the crosshairs. The India AI Impact Summit 2026 wasn’t another safety-focused talkfest — it was a power grab, and India just pulled it off. The Numbers Are Staggering Reliance Industries led with $109.8 billion over seven years for AI and data infrastructure across India. The Adani Group followed with $100 billion by 2035 for renewable energy-powered AI data centers. Microsoft confirmed $50 billion by 2030 for the Global South, with India as the flagship market. Yotta Data Services announced $2 billion-plus for one of Asia’s largest AI computing hubs running NVIDIA’s Blackwell Ultra chips. ...

Stop watching Silicon Valley for a second. The most important AI event this week is happening in New Delhi — and the chaos tells you everything about where India actually stands in the global AI race. Every AI CEO on Earth Just Flew to Delhi Sam Altman. Sundar Pichai. Dario Amodei. Demis Hassabis. Yann LeCun. Brad Smith. Over 100 countries. Twenty heads of state. 3,250 speakers. 250,000 expected visitors across five days at Delhi’s Bharat Mandapam. ...