Skip to content
Advertisement
All-In Podcast’s David Friedberg slams Trump plan to force disclosure of Chinese AI use

All-In Podcast’s David Friedberg slams Trump plan to force disclosure of Chinese AI use

The co-host called restrictions on open-source AI a path toward an Orwellian state, two days after Reuters reported the proposal

Washington wants to know which American companies are running Chinese AI models under the hood. At least one prominent Silicon Valley voice thinks asking the question is the problem.

On Friday’s episode of the All-In Podcast, co-host David Friedberg tore into a Trump administration proposal that would require US companies to disclose their use of Chinese AI models. He framed it as the first step on a slippery slope for open-source software.

What Friedberg said, and what the proposal does

Friedberg did not hedge. He argued that any attempt to curb, restrict or even label open-source AI would shrink both human prosperity and individual liberty.

“I think this continues to be the path of tyranny,” Friedberg said.

He went further, warning that the end result could be what he called an Orwellian state. His prescription was the opposite of regulation: open-source AI should be allowed and actively encouraged to spread.

His reasoning is that without open alternatives, power concentrates in fewer hands. In Friedberg’s telling, the groups left standing would be tyrannical ones, or regulated monopolies that face little competition.

To make the case, he pointed to Apache and Firefox. Both are open-source projects that helped keep core internet technology accessible rather than locked behind a single gatekeeper.

Advertisement

The episode aired on November 20, 2026. That was two days after Reuters reported on November 18 that the administration is weighing the disclosure rule.

The scope matters here. The measure under consideration would make companies publicly disclose their use of Chinese AI models, but it would not ban those models or impose heavy restrictions on them.

No formal rulemaking timeline has been confirmed. Enforcement details have not been spelled out either, so the proposal remains exactly that.

Why Chinese models are on Washington’s radar

Airbnb has been cited as a user of Alibaba’s Qwen model in discussions around the disclosure plan.

Congress is already poking around. The House Select Committee on China has sent inquiries to companies including Airbnb and DoorDash about their use of Chinese AI models.

That context explains why a disclosure rule is surfacing now. Lawmakers have been asking individual companies one letter at a time, and a rule would effectively make the answer mandatory for everyone.

Friedberg’s argument flips that logic. If open code is the thing being scrutinized, he suggests, the real winners are the closed, well-connected incumbents that never have to publish anything.

A crowded episode for AI news

The disclosure debate was not the only AI story on the hosts’ plate. The same episode also covered Anthropic’s IPO and the stock market challenges facing Nvidia.

What this means for companies and investors

For now, the practical impact is limited, because there is no rule. There is a reported proposal, a congressional inquiry campaign and a lot of loud opinions.

The second is how companies already named in congressional inquiries, such as Airbnb and DoorDash, respond publicly. Their answers could become a template for the rest of corporate America.

The third is whether the definition of “Chinese AI model” stays narrow or expands. A rule that captures fine-tuned derivatives or models built on Chinese base weights would reach far more companies than one focused on direct use.

Friedberg is betting that openness is the better defense. Washington, at least in the version Reuters described, appears to be betting that visibility comes first, and that the market can sort out the rest once everyone has shown their cards.

Disclosure: This article was edited by John Chen. For more information on how we create and review content, see our Editorial Policy.