ALFA Field Notes

If Tariffs are the Play, AI Export Controls Undermine from Within

President Trump has long promised tariffs, and now they’re here. In our book, this week’s actions were a clear “known known” and should not have taken anyone by surprise.

But whether you love tariffs or hate them, they will not operate in a vacuum as the Administration works towards its stated goal of rebuilding America’s industrial base and maintaining our technological edge. It is essential that policymakers in the White House and on Capitol Hill also create the right tax and regulatory environment to bolster this new economic framework.

For one, that means keeping taxes low–not raising them on individuals, investments, and businesses, as recent reports have suggested are currently under discussion for the upcoming reconciliation package.

Likewise, it means rolling back regulations that make it harder to build in America and harder to sell American goods to the world. On that front, there is perhaps no rule as counterproductive as the Biden-era AI diffusion rule, restricting one of America’s most competitive export sectors—advanced AI chips—at precisely the wrong time.

Rather than boosting our position, this rule slows down the very export growth we need. With global AI infrastructure spending projected to hit $1 trillion by 2026, U.S. chip dominance is arguably our single biggest advantage in closing trade deficits.

Yet just as our trading partners are being pressured to open their markets to American goods, we’re simultaneously tying our own hands in the sector where we presently have a clear lead and where other countries are actively seeking out our technology. Meanwhile, DeepSeek’s recent breakthrough on Huawei chips proves the rule’s fatal flaw: it doesn’t stop foreign AI progress. It merely gives companies like the CCP-controlled Huawei an open lane to seize crucial market share in our absence.

If we’re serious about fixing global trade imbalances, revoking the AI diffusion rule is the next logical move. Otherwise, we’re just making things harder on ourselves and weakening our strongest export for no good reason.

"A Game Changer": Pennsylvania’s Bid to Become the AI Infrastructure Hub

In Washington, we are often consumed by the macro. Perhaps that’s due to the disappearance of local news outlets or the rapid-fire nature of online discourse today. But it’s not the marco that resonates. It’s the micro.

When we talk about energy and AI infrastructure, yes, meeting global demand and competing with China are crucial. But just as important is understanding what these massive projects mean for the communities where they take shape. How do the people living there actually feel about them?

Thankfully, journalists like Salena Zito have dedicated their careers to telling those local stories. So when news broke this week that the largest gas-powered plant in the U.S. would replace Pennsylvania’s largest coal-fired plant to power a massive data center campus, Salena was on the ground in Homer City, Pennsylvania capturing the community reaction.

It’s a technology, economic, and cultural story.

“A game changer for the region and for the state of Pennsylvania,” is how business agent for the International Brotherhood of Boilermakers Local 154, Shawn Steffee described it to Salena.

He continued: “This is great for my hometown [and] my school district that my family has gone to for six generations of kids. My family still lives there. I still live here in the community, and this is definitely a shot in the arm for a county that’s been really economically challenged.”

As a technology policy group, we celebrate innovation. But in the political world—where your ALFA authors spent over a decade of their lives—we know that progress also requires public buy-in. The Homer City project is proof that American workers and communities aren’t just tolerating the future. They’re embracing it.

Freedom Clusters Are Here

In February we made the pitch for data center development on federal land — what we called Freedom Clusters. Yesterday, the Department of Energy began the process to leverage some federal land to support the growing AI infrastructure demand.

From the announcement: “DOE has identified 16 potential sites uniquely positioned for rapid data center construction, including in-place energy infrastructure with the ability to fast-track permitting for new energy generation such as nuclear”.

This is a critical move at a crucial time. At the hearing The Economics of AI, Data Centers, and Power Consumption, Executive Director of the National Center for Energy Analytics Mark Mills testified: “The amount of energy used to launch a rocket is consumed every day by just one AI-infused data center” while the Abundance Institute’s Neil Chilson warned: “the energy regulatory environment creates crippling bottlenecks that threaten AI dominance.”

By DOE going this route, we will be able to push against traditional limits slowing down infrastructure deployment, specifically with new nuclear energy — which is what Oklo CEO Jacob DeWitte predicted in a recent Bloomberg interview.

We love to see it and will keep you posted on the RIF progress.