The Trump administration is locked in a secret internal war over how to handle China's artificial intelligence boom

 **

The Trump administration is locked in a secret internal war over how to handle China's artificial intelligence boom** — a battle that pits national security hawks pushing for an all-out ban against free-market advocates warning that such a move would cede America's technological edge.

At the center of the storm is **Moonshot AI's Kimi K3**, a cutting-edge Chinese model that is nearly as good as American rivals but costs a fraction of the price. Its release last week has reignited efforts to choke off Chinese AI, exposing deep fractures within the White House over how — and whether — to fight back.

---

 The Secret Offensive: A Year of Quiet Maneuvers

For more than a year, parts of the Trump administration have been quietly waging a campaign to cut off U.S. access to Chinese AI, according to knowledgeable sources who spoke to Axios. The efforts have included:

- **Adding Chinese AI labs to the Commerce Department's "Entity List,"** which would effectively cut off U.S. access without a license.

- **A National Security Agency advisory** warning U.S. companies about the threats posed by Chinese AI labs.

- **A proposed executive order** requiring U.S. companies to guarantee security and accept liability if they hosted Chinese models.

- **Draft rules from the Commerce Department** leveraging supply-chain authorities to target Chinese open-source models.

All of these efforts were killed by administration officials who feared that heavy-handed regulation would stifle American innovation.

---

## The Shift: Hawks Ascendant

The calculus changed when **Sriram Krishnan**, a key White House adviser who opposed a ban, left his position. His departure — combined with the rise of more powerful Chinese models and fresh cybersecurity fears — has given national security hawks the upper hand.

"The rise of Kimi is reigniting those efforts," a source told Axios.

The shift in personnel has been pivotal. With pro-competition voices diminished, the momentum for a crackdown is picking up again. The administration is now signaling that it could ban cutting-edge Chinese AI models — a move that would lock in dominance by **OpenAI and Anthropic**.

---

## The Split: Free-Market vs. National Security Hawks

The internal battle is defined by two competing visions:

**The National Security Camp** argues that Chinese AI models pose an unacceptable risk. The Commerce Department, NSA, and White House Office of the National Cyber Director have all weighed measures to discourage U.S. companies from using Chinese technology. Their approach includes procurement rules, Entity List threats, and public pressure campaigns aimed at companies that rely on Chinese models.

Instead of an outright ban, some officials are pushing a **"slower and more durable"** strategy — highlighting potential backdoors and security vulnerabilities in Chinese models. "It's an offensive approach where the administration encourages a more innovative U.S. open-source ecosystem," a source familiar with government discussions said.

**The Free-Market Camp** warns that a ban would backfire spectacularly. **David Sacks**, an outside White House AI adviser and former AI and crypto czar, has been the most vocal critic. He wrote Sunday on X:

> **"We are at a critical inflection point in AI policy. The leading closed labs, already a duopoly in terms of AI model revenue, want the government to eliminate their open-source competition."**

Sacks has long warned against **"regulatory capture"** — the danger that OpenAI and Anthropic are using national security concerns to eliminate cheaper, open-source rivals. He argues that "permissionless innovation" is what won America the internet era, and that overregulation is how the U.S. loses the AI race.

> **"If we tie ourselves in knots, the rest of the world isn't going to play by our rules,"** Sacks said.

---

## The Kimi K3 Catalyst: A Chinese Breakthrough

The release of **Kimi K3** has transformed the debate from abstract to urgent. The model is the first to top Arena AI's front-end code ranking, performing at or near the frontier of U.S. capabilities. Its 2.8 trillion parameters and 1 million-token context window put it in the same league as Anthropic's Fable and OpenAI's ChatGPT — but at a fraction of the cost.

U.S. companies are increasingly turning to Chinese open-source models because they're cheaper and, as Kimi demonstrates, nearly as good. That trend has alarmed national security officials, who fear that American dependence on Chinese AI could erode the U.S. technological advantage.

---

## What Comes Next: A Ban or Something Else?

The administration faces a series of difficult choices:

**An outright ban** would be momentous — and almost certainly challenged by tech companies that rely on affordable open-source models. But it would also cement OpenAI and Anthropic's dominance.

**A "chilling effect" strategy** is more likely, sources say. This would involve procurement rules, Entity List threats, and public pressure campaigns to make Chinese models too risky for U.S. companies to use. Leading AI labs are reportedly approaching the administration every **3-5 months** with ideas to ban open-source models.

The bottom line, according to Sacks, is that the largest U.S. labs have **"laid their cards on the table"** — they want the government to eliminate their open-source competition. The question is whether the rest of Silicon Valley — "the vast majority that still values open competition" — will fight back.

---

## Frequently Asked Questions

**Q: Why is the Trump administration considering banning Chinese AI models?**

A: National security officials fear that U.S. companies are becoming dependent on cheaper, nearly-as-good Chinese open-source models, which could erode America's technological advantage and pose security risks.

**Q: What is Kimi K3 and why does it matter?**

A: Kimi K3 is a Chinese AI model released by Moonshot that performs at or near the level of top U.S. models but costs far less. Its release has reignited concerns about Chinese competition.

**Q: Who is David Sacks and what is his position?**

A: Sacks is an outside White House AI adviser and former AI and crypto czar. He warns against "regulatory capture" — the danger that OpenAI and Anthropic are using national security concerns to eliminate open-source competition.

**Q: What is the "regulatory capture" argument?**

A: The argument is that dominant U.S. AI labs are pushing for regulations that would eliminate cheaper, open-source rivals — using national security as a cover to protect their market position.

**Q: Will the U.S. ban Chinese AI models?**

A: It's unclear. Some officials are pushing for an outright ban, while others favor a "chilling effect" strategy using procurement rules and public pressure campaigns. No final decision has been made.

--Read more-

The secret battle inside the Trump administration is about more than just policy — it's a fight over the very nature of American innovation. Will the U.S. win the AI race through open competition, or will it try to wall itself off from the rest of the world? The answer will shape not just the AI industry, but the future of the global economy.