America's Real AI Rival Isn't Beijing — It's Washington's Own Backyard

.

A Warning From Inside the AI Industry

Shyam Sankar, chief technology officer of the U.S. software firm Palantir, says America's biggest problem in the AI race isn't foreign competition. It's domestic gridlock over power and infrastructure. He made the comments in an interview at Palantir's American Builder Summit, an event built around showcasing how ordinary workers already use AI on the job.

Palantir builds data and AI software for the Pentagon, U.S. intelligence agencies and large manufacturers. Sankar argued that the country risks slowing itself down more than any rival could. He summed it up sharply: the only real threat to American AI-driven prosperity is self-inflicted, not caused by an outside enemy.

Data Centers Spark a Political Backlash

The United States and China are both investing heavily to lead in artificial intelligence, since the technology is seen as key to economic growth, military strength and global influence. But that race depends on building huge numbers of data centers (large facilities full of computer servers) along with the electricity supply to run them.

That buildout has run into fierce local opposition. Communities have raised concerns about rising electricity bills, heavy water use, noise, traffic and who ultimately pays for grid upgrades. Some projects have already been delayed or scrapped, and several state legislatures have proposed limits or outright bans.

The most striking example: New York became the first U.S. state to freeze new large-scale data center construction, when Governor Kathy Hochul signed a one-year moratorium on facilities using 50 megawatts of power or more. The governor described it as a way to build the country's strongest regulatory standards for data centers while giving local communities more support. Polling has shown many Americans are uneasy about the pace of construction, even as they want the economic benefits AI promises.

President Trump publicly criticized the New York freeze, arguing that data centers are a major source of future jobs and investment for the states that host them. The episode illustrates how the data center question has become a live political fight, not just a technical one, and it has already spilled into 2026 midterm campaigning in states like Pennsylvania.

Proof From the Factory Floor

Sankar's answer to the backlash was not more debate, but concrete examples. He described a manufacturer of submarine parts that cut its production-planning time from two weeks down to less than an hour using AI tools — a change that freed up enough capacity to add an entire extra work shift.

He noted that most workers on that new shift have no idea artificial intelligence made their jobs possible in the first place. For Sankar, that is the point: AI is already quietly making American workers more productive, not replacing them.

That message shaped the whole summit. Instead of unveiling flashy new AI models, Palantir put factory employees, nurses and military personnel on stage to demonstrate tools they had built themselves to solve real problems at work.

Chinese Models Built on "Stolen" American Work

Beyond the domestic power fight, Sankar has separately raised a second concern about China: that some of its newest AI systems were built by copying, rather than developing, U.S. technology. He told Bloomberg Television that a wave of new Chinese open-source models looks like the result of so-called "distillation" — training a cheaper model to imitate a more powerful one by studying its output — and argued that much of this work amounts to lifting American research (calling it "stolen" intellectual property).

His claim fits a broader pattern. U.S. AI company Anthropic has separately accused a major Chinese technology firm of running a large-scale distillation campaign against its own AI model using thousands of fake accounts. The U.S. Treasury Secretary has even floated the possibility of sanctions against Chinese AI firms found to have stolen American technology.

Not everyone agrees on how serious the problem is. Some AI industry figures, including a leading open-source executive, have argued distillation is a common technique used by companies everywhere, including in the United States, and that China's progress owes more to strong research teams than to copying. Microsoft's CEO has also pointed out an irony: large AI labs that train freely on public data have sometimes complained when others study their own models in a similar way.

What Happens Next

Sankar argues the U.S. is stuck chasing a distant, futuristic vision of "artificial general intelligence" (a theoretical system that could match or beat humans across almost any task) while ignoring the gains today's, already-existing AI could deliver right now. He compares the moment to the early days of electricity: the technology existed, but American industry hadn't yet reorganized around it.

For him, the fix isn't more caution — it's faster deployment, paired with serious national investment in the power grid needed to support it. Whether that view wins out will depend heavily on politics at the state level, where lawmakers are increasingly caught between AI's economic promise and voters' anger over rising utility bills.

The debate is unlikely to cool down soon. With more states weighing their own restrictions and the Trump administration pushing hard the other way, the fight over data centers may end up shaping America's AI future as much as anything happening in Beijing.

.

Sources

1. Reuters (via Yahoo News) — "New York becomes the first state to impose a data center moratorium": https://www.yahoo.com/news/politics/articles/york-becomes-first-state-impose-090455587.html
2. The Washington Post — "New York becomes first state to impose data center moratorium": https://www.washingtonpost.com/technology/2026/07/14/new-york-becomes-first-state-impose-data-center-moratorium/
3. CNBC — "Trump blasts New York Gov. Hochul over AI data center moratorium": https://www.cnbc.com/2026/07/15/trump-blasts-new-york-gov-hochul-over-ai-data-center-moratorium.html
4. Bloomberg News — "Palantir's CTO Sees Chinese AI Models Posing Economic Risk to US": https://www.bloomberg.com/news/articles/2026-07-15/palantir-s-cto-sees-chinese-ai-models-posing-economic-risk-to-us
5. Governor Kathy Hochul's Press Office (New York State) — "First Statewide Moratorium on New Hyperscale Data Centers": https://www.governor.ny.gov/news/first-statewide-moratorium-new-hyperscale-data-centers-launched-governor-kathy-hochul
6. TechCrunch — "US threatens sanctions against Chinese AI models over IP theft": https://techcrunch.com/2026/07/21/us-threatens-sanctions-against-chinese-ai-models-over-ip-theft/

.