The Great American Tech “Crackdown” Is Looking Like China’s

If I had asked you to guess which country’s tech company has to ask its government for an approved list of customers before releasing a new product, you likely would have guessed China or maybe Russia.

Except today that honor belongs to OpenAI in America.

If I had asked you which country’s government gets “golden shares” in companies it deems strategic or wants to keep a closer eye on, your answer might be China or even Germany.

Both answers would be correct, except America is trying to join that rank of socialistic best practices, especially when it comes to AI.

And if I had asked you to recall which entrepreneur made high profile speeches and actively engaged with his government to influence the way it regulates his company, but only to get more than he wished for, Jack Ma might come to mind first.

He indeed made such a speech telling the Chinese government how it should regulate the fintech industry, which I translated almost six years ago. It’s safe to say he got more regulation than he asked for, but not the kind he wanted. If you swap the word “speech” for “essay” or “blog post”, then Dario Amodei’s “Policy on the AI Exponential” post would be just as good of an answer. The US government banned access to the Fable model by any non-US citizen anywhere in the world, two days after Amodei published his essay; probably not the kind of regulation he wanted either.

Jack’s speech put the Ant IPO on ice indefinitely. Dario’s essay may not have thrown a wrench into Anthropic’s IPO yet, but its rival yet also comrade in enduring the arbitrary and opaque current model approval process, OpenAI, might delay its public listing.

No analogy is a 100% match (otherwise it would be called a simile). But what is happening today to frontier AI labs in the US – the Great American Tech “Crackdown” – bears many similarities to what happened to China’s tech sector five years ago.

I’m putting the “crackdown” in quotes, of course, because things are not that bad yet, but if we don’t start talking about the warning signs and the rather dangerous direction of travel, it very well could be.

Americans have not been good at learning from what China is good at. That requires humility we aren’t born with, patience we don’t reward, and the setting aside of pride that runs counter to our culture. If anything, pride will be on its most grand and raw display yet, as we approach the 250th birthday of this young nation next week.

Perhaps, we can at least learn from what China did that wasn’t so good and avoid the same fate. China’s tech crackdown would be one of them.

Since Jack’s impassioned but ill-fated speech, multiple tech sectors either disappeared (education) or were permanently hobbled (gaming) overnight. Entrepreneurs and investors would never make a move or write a big check unless they are sure their actions don’t run counter to the government’s preferences. The ones who can’t live and thrive under this new environment would either move abroad, giving life to terms like “Singapore-washing”, or build in China, where talent is still strong and affordable, but only serve customers and users overseas.

Manus is Exhibit A in checking both boxes. That didn’t end so well. America is still far from producing its own Manus tragedy. But the slope of overreach is slippery and moves fast.

Whatever preconceived notions or stereotypes you may have about how the Chinese or the American system works, those notions are being flipped on their heads, especially when it comes to AI. I’m not the only one noticing this.

Marc Andreessen said the following on a podcast with a Washington think tank:

“It is really remarkable that China has decided that open-source AI is something that is good and that they want to exist, and they want to propagate…We’re in a weird state of the world where the supposedly totalitarian regime is trying to open up the technology, and the supposedly democratic government system is trying to restrict and control technology

The word “supposedly” is doing all the work here. While China is promoting open source AI (the country has a long history with open source which I explored previously), America may be closer to banning it than ever before, while bureaucrats fine-comb the prospective customers list of OpenAI and Anthropic’s CRM.

There are many things that I wish America and China could be more similar in. America could use some newer infrastructure, stronger public safety, and industrial policy planning that lasts more than four years. China could use more pluralism, rule of law, and a national soccer team that would advance to the World Cup knockout round.

But imitating each other’s tech crackdowns should not be one of them.