Friday, July 10 July 10, 2026

Good morning. It's Friday, July 10th, 2026, and today the AI story is geopolitical — and it's moving fast.

Let's start with the biggest headline. Beijing has ordered Meta to unwind its two-billion-dollar acquisition of Manus, the AI agent startup. That's according to reporting from the Financial Times this morning. And Tencent is already moving in — the Chinese tech giant is now in advanced talks to become Manus's largest shareholder. This is a direct, government-forced reversal of a major U.S. tech acquisition, and it sends a clear signal: China is treating its homegrown AI companies as strategic assets, not products for sale to Silicon Valley.

But that's only half the story. Reuters is reporting — exclusively — that Beijing is now considering restricting overseas access to China's most advanced AI models entirely. We're talking DeepSeek, Alibaba's Qwen, ByteDance's models — the same tools that shocked the world earlier this year with their performance at a fraction of Western costs. Chinese authorities have held meetings with top tech firms over the past month. The move mirrors what the U.S. has already done with export controls on chips and frontier AI. Both superpowers are now pulling up the drawbridge on their most powerful AI.

Here's what makes this even more complex. The Financial Times also reported today that OpenAI and Google have been selling AI services to Singapore-based subsidiaries of Alibaba, Baidu, and Tencent — entities that are on U.S. blacklists. So while Washington restricts chip exports to China, American AI companies are quietly selling model access through offshore workarounds. Expect this loophole to become a major regulatory flashpoint in the coming weeks.

On the money side, Bloomberg data out this week shows global M&A has hit 3.2 trillion dollars so far in 2026 — one of the biggest dealmaking years on record — and AI is the primary engine. Companies are acquiring AI capabilities, infrastructure, and talent at a pace we haven't seen since the dot-com era. For context, that 3.2 trillion figure is fueled by an ebullient stock market, massive AI bets, and an open regulatory environment — at least in the U.S.

That regulatory environment may not stay open much longer. AI Giants are pouring millions into rival PACs as Congress drafts AI legislation. Two competing industry coalitions are lobbying for incompatible frameworks right now — one pushing for federal preemption of state laws, the other pushing for stronger liability rules. The lobbying battle is heating up just as the geopolitical pressure builds.

One more funding note worth watching: Oxylabs, a web data scraping infrastructure startup, just raised 130 million dollars from Warburg Pincus, hitting unicorn status. In a world where AI models need massive amounts of fresh web data, the picks-and-shovels play is web data infrastructure — and investors are noticing.

And in the mainstream discourse, AI meeting notetakers are drawing scrutiny. ABC News is running a piece today about professionals pushing back on tools like Otter and Fireflies over privacy concerns — specifically, the discomfort of having every conversation summarized and stored. This is the adoption friction story that enterprise AI vendors are going to have to solve.

So here's the throughline for today: the AI Cold War is no longer metaphorical. China is using regulatory force to keep its AI assets out of American hands, restrict foreign access to its frontier models, and consolidate ownership under domestic tech giants. Meanwhile, American companies are finding workarounds — and Congress is watching. The next few months will likely produce the most consequential AI governance moves we've seen yet.

One business idea to close on: with both the U.S. and China restricting access to their frontier AI models, there's a real opportunity for a neutral-country AI model brokerage — a compliance-aware platform that helps enterprises legally access and deploy the right model for their jurisdiction. Think of it as a regulatory-native AI API layer, built for the new world of AI export controls. The companies that solve cross-border AI compliance first will have a serious moat.

That's your Morning AI briefing for July 10th. Stay sharp out there.