Representative Ron Estes (R‑KS) recently introduced House Resolution 1340, opposing foreign digital services taxes (DSTs). It is co-led by Rep. Suzan DelBene (D‑WA) and cosponsored by members of both parties. The resolution comes as the EU again weighs a bloc-wide digital levy of its own, giving political cover to member states to keep, and even expand, their national taxes. 

The resolution sets out a number of important principles that DSTs violate: net taxation rather than gross taxation, no double taxation, legal certainty, and taxing rights tied to physical presence.

Opposition to DSTs has been a recurring feature of congressional bipartisanship, and it is shared across administrations. Trump’s trade office opened the first investigation in 2019, which the Biden administration continued. In early 2025, Trump issued an order to renew investigations into a half-dozen additional European DSTs. 

Washington’s consensus centers on the discriminatory mechanics of DSTs. They tax gross revenue instead of profit, so it applies even if the company is losing money in the jurisdiction. It reaches firms with no physical presence in the country levying the tax, and the tax is typically designed to catch American platforms while leaving domestic firms untouched.

DSTs are interesting because they are most strongly opposed by American politicians. But their costs fall mostly on consumers. This point came up repeatedly at a Cato webinar I hosted last month with European tax economists. The DST is assessed and remitted by large digital platforms, but the economic incidence falls on consumers in the jurisdiction imposing it.

Digital companies have been transparent about how they handle these levies. Google adds jurisdiction-specific surcharges to advertisers’ invoices that closely follow local DST rates. Apple adjusted developer proceeds in the UK, France, and Italy to reflect DST costs and raised app prices in countries like Turkey, where the combined tax burden was large enough to warrant a price change. Meta had been the holdout, absorbing the cost of DSTs for years. That changes on July 1, 2026, when Meta introduces “location fees” to pass DST costs directly to advertisers across several European countries. 

A recent study confirms the same pattern for Amazon, which now charges a standalone Digital Services Fee in countries where it faces a DST. Economists Dominika Langenmayr and Rohit Reddy Muddasani find that third-party sellers pass those higher fees on to consumers. In France, Spain, and the UK, they find that consumers ended up bearing between €/£1.50 and €/£2.50 for every euro or pound the tax actually raised. They find Italy is an exception. 

American politicians are still right to push back against these taxes. A tax’s economic damage routinely exceeds the revenue it raises. So even when the cost is fully passed through to consumers, US firms still lose through forgone sales and thinner margins. This is especially true when the tax explicitly exempts their foreign market competitors. 

The bigger problem, and the one at the core of the House resolution, is that DSTs erode the international income tax system that has shielded cross-border trade and multinational firms from punitive and duplicative taxes for more than half a century. DSTs are not the only threat to that order. The return of tariffs is a direct threat. The Organisation for Economic Co-operation and Development (OECD) has also spent the better part of the last decade diligently undermining the protections of physical presence and rules against double taxation.

The Rep. Estes resolution’s principles are good: net over gross taxation, no double taxation, certainty, and taxing rights tied to physical presence. They should apply as broadly as possible. The same principles that condemn European DSTs, impugn tariffs, the US’s multiple overlapping corporate minimum taxes, and emerging proposals to tax AI.

Congress should oppose bad taxes wherever they appear. The principles behind Washington’s consensus against DSTs are a good starting point. 

True liberty cannot exist without economic and financial autonomy. Every individual, regardless of their background, must have the right to protect their wealth. They must have the right to access open markets by transacting freely, without the shadow of state corporatism or financial surveillance. 

Central Bank Digital Currencies (CBDCs) are a direct threat to these rights. 

They are money that can be programmed by the government. They can be turned off completely or just for spending on items the state disapproves of. They are a surveillance-punishment system dressed in the language of financial innovation. They are not “just another digital form” of money.

That is why my Cato colleagues and I have worked so hard to get the truth out about CBDC’s.

Last week, my colleague Nick Anthony was invited to Prague to talk about his work and the danger of CBDC’s. Here’s an excerpt from the speech that he gave:

This is not a distant threat. It is not a story about authoritarian governments in faraway places. It is a story about the architecture of money itself—and who controls it. That affects all of us.

When money becomes programmable, the question is no longer just “How much do you have?” It becomes “What are you allowed to do with it?” That is different from the power to tax or the power to inflate. It’s the power to permit.

The rest is available here and is worth checking out. 

As we prepare to commemorate Juneteenth, it’s also worth remembering that liberty has historically been undermined not just by social codes but also by institutional barriers that locked vulnerable populations out of commerce and wealth creation. These kinds of top-down financial systems have always placed the heaviest burdens on those with the least economic power, and they always will.

That’s why CBDCs are not compatible with freedom.

With July 4th looming, it’s tragically poetic that the country is again facing the question of how to constrain the exercise of arbitrary power. The Trump administration’s decision — based on questionable justifications — to block foreign nationals from using two Anthropic models indicates that AI governance is being shaped by actors who wield incredible influence over this critical domain and are yet subject to few effective constraints. This episode strengthens the case for Congress to step up and pass an AI governance framework that will reduce the frequency with which AI policy is breaking news.

What Happened

A brief review of the known facts (of which there are not many—itself a troubling sign from a rule of law perspective) surfaces the many ways in which highly consequential AI policy decisions seem disconnected from a reasoned, transparent, and thorough process. Commerce Secretary Howard Lutnick sent Anthropic a letter on Friday detailing that their latest models—Fable 5 and Mythos—are now subject to export restrictions. 

Mythos drew popular attention following documented instances of its being able to detect and exploit cyber vulnerabilities. Anthropic decided to share that model with a select set of trusted partners rather than risk bad actors leveraging the technology for nefarious ends. Fable 5 is a version of Mythos with several additional safeguards and was therefore made popularly available. In fact, Anthropic reports that Fable 5’s safeguards are “substantially more effective than those of any previously deployed model.”

The administration’s imposition of an export restriction followed from evidence of a potential narrow, non-universal jailbreak of Fable 5. Amazon researchers probing the model determined that it could provide users with information about cyber vulnerabilities when directed at websites. Yet, as Katie Moussouris, chief executive with the cybersecurity firm Luta Security, told the Wall Street Journal, such information is far more valuable to defenders of Internet systems than attackers. 

Nevertheless, the administration’s directive means that Anthropic can no longer make those models available to foreign nationals, theoretically including their own employees. Anthropic responded by blocking everyone’s access to Fable 5 and Mythos to ensure it does not violate the administration’s expansive directive.

A Failure to Adhere to the Rule of Law

In conjunction with my frequent co-author Alan Rozenshtein, I have previously reported on the ways in which AI governance by ad hoc executive action (or, in many cases, inaction) and by exclusively private efforts raise rule-of-law concerns. The rule of law can be defined in many ways by myriad actors, but most would agree that its core tenets include publicity, prospectivity, and legality. Publicity mandates that the affected party know the rules and laws that govern their conduct. Prospectivity ensures that the affected party is able to align their behavior with any applicable rules and laws, rather than be held accountable for prior acts. Legality confines every actor to only operate where they have clear authority to do so. 

Failure to adhere to those tenets undermines the extent to which individuals can live an ordered life and monitor those with extensive influence over their well-being. Government officials opting not to adhere to the rule of law also imposes significant costs on our economy. “The rule of law is the most influential factor for long-term economic growth and societal wellbeing,” as Annie (Yu-Lin) Lee and Joseph Lemoine summarized in an Atlantic Council report. Perpetuating an AI governance approach in which enforcement decisions are publicly announced on a Friday during the middle of a World Cup game is unlikely to safeguard individual liberty or encourage innovation and entrepreneurialism. 

That’s precisely why the administration’s recent action requires scrutiny. And that’s precisely why Congress must act so that such consequential governance decisions are not disclosed via X posts and are instead made in alignment with the Rule of Law. 

On publicity, the administration’s decision to take extraordinary action against Anthropic was disclosed by Anthropic—not the administration. Of course, the government necessarily withholds information or delays public announcements of enforcement actions in many cases. When it comes to decisions that may deny millions of people access to AI technology that is already shaping our economy, such opacity is in immediate conflict with the rule of law.

On prospectivity, a scroll of X makes clear that no one saw this coming. That’s not a good thing. While some AI scholars have flagged that export control authorities may one day be used to block individuals from accessing certain AI models, it has not been the topic of popular discussion. In turn, it has not been a potentiality actively shaping the decisions of AI labs. This surprising turn to export restrictions—even after Anthropic and other labs have engaged with the administration on numerous occasions about the cyber risks posed by their models and the safeguards taken to mitigate those concerns—instead suggests that labs will have to constantly be looking over their shoulder for what ‘creative’ enforcement decisions may come down the pike. 

On legality, research on the export restrictions raises significant doubts about whether these provisions are intended to reach AI models. As Joe Khawam and Tim Schnabel wrote in Just Security, “[T]hese frameworks, designed for discrete transfers of static information between known parties, are ill-suited to govern AI systems that generate unlimited, dynamic outputs on demand for potentially anonymous users worldwide.” Perhaps the administration has a clearer explanation as to why the legal authority it’s relying on in this case fits the facts. However, that explanation has not yet been made available. 

What has been made clear is that if the administration thinks such restrictions are legally available and necessary, it ought to limit access to many more such models offered by other labs. Anthropic is not alone in offering models with advanced cyber capabilities, which were purportedly the basis of the administration’s decision, as former AI czar David Sacks spelled out on X. The related rule-of-law concept of generality mandates that like cases be treated alike, which implies that the administration should be in the process of sending letters to other labs.

Conclusion

This is the second time in as many weeks that I’ve written about wild swings in AI policy. It should not be the case that the government needs to be reminded by private entities about the rule of law. Yet, here’s what Anthropic had to say on Friday: “[W]e believe the government should have the ability to block unsafe deployments as part of a statutory process that is transparent, fair, clear, and grounded in technical facts. This action does not adhere to those principles.” 

Transparency, fairness, clarity, and evidence-driven enforcement should be core components of AI governance. So long as AI policy decisions are driven by a few actors taking action behind closed doors in response to non-public reports, we will be far short of that standard.

In 2021, Farryn Slaton-Barkley picked up her young son from daycare and noticed something was wrong. He kept repeating his teacher’s name, which was unusual for a child who was low verbal and usually only communicated his basic needs. Then she saw the marks on his hands.

An investigation by the state licensing agency confirmed abuse. Farryn, a veteran teacher with a PhD who has taught in public schools and as an assistant professor at both public and private universities, blamed the childcare system rather than just the worker. So she decided to channel her outrage in a way that would make lasting changes. 

Using a grant from a local violence prevention initiative, Farryn trained women in Macon, Georgia, to open their own daycares, equipping them to handle children of all neurotypes. That was the seed of what would eventually grow into Rising Scholars Collective.

The organization’s work is wide-ranging, but the thread running through all of it is expanding academic options for neurodivergent children. Farryn describes Rising Scholars as “really an advocacy company” whose “main job is to help systems and educational organizations be more manageable, be more inclusive for children with disabilities.” 

One of her key programs sends twice-exceptional middle schoolers—kids who are gifted but also have a disability—to college in the summer. She launched the first cohort in 2023, and this summer she’s opening it up to the broader community. The goal isn’t just academic acceleration. It’s making sure kids who can handle college-level coursework still get to be around peers their own age. “They might be able to do calculus, but you know, they might need help with writing,” Farryn explains. “They might still need to help tie their shoes. So just making sure they’re with their age group is important at this point in their development.”

Rising Scholars also helped organize the Middle Georgia School Choice Fair, which grew from eight vendors and roughly 200 attendees in its first year to 48 registered vendors and nearly 400 attendees in year two. It’s now formally partnering with the National School Choice Foundation, and Farryn is already planning for 2027.

Another big focus is growing the microschool ecosystem in an underserved community. Through what she calls the Microschool Builder Series, Farryn connects aspiring school founders to funding sources and national networks. One microschool has already launched out of her trainings. Another is being planned by a respite provider who wants to serve nonverbal, high-needs autistic children who are currently left with few alternatives to public placement. “You have parents take their kids out of institutions for that and homeschool them because it’s not safe,” Farryn says. “They’re not made to feel safe. Just like what happened to my son when he was younger, it wasn’t safe.”

She also works one-on-one with homeschool families, helping design individualized learning plans, connecting students to summer programs, and assisting families in applying for Georgia’s Promise Scholarship and Special Needs Scholarship. As an example, one of the families she serves includes a teenage girl with special needs who is interested in photography. Farryn helped enroll her in a college and career charter program so she could take film and trade classes with peers. “She’s just glowing,” Farryn says. “It is really about looking at the student, finding the needs, finding the programming, getting them enrolled, and then just seeing them flourish.” 

Farryn is still in the classroom herself, teaching high school. “That informs my practice because I am a parent instructor, so I see things from a parent point of view,” she says. Her own son attended four different schools before landing in a microschool that met his needs. In her classroom, she encounters students who’ve masked their neurodivergence for years, picking up coping mechanisms and trauma along the way, with no diagnosis and no support. She uses April’s Autism Awareness Month as an entry point to talk about regulation, neurotypes, and brain differences, giving students language they can use when they eventually access health services on their own.

Rising Scholars Collective is not Farryn’s full-time job. “It is really a labor of love because it’s necessary. And our children, whether they do private, whether they do public, whether they do virtual, are just always at such risk,” she says. “It’s just like the same story over and over of how they’re underserved, not supported, you know, and, like what happened to my son, physically assaulted.” Rather than sit back and complain, she decided to build a stronger community and empower parents through advocacy, homeschool and microschool coaching, and school choice fairs.

“The irony is that I exist because the system’s broken,” she says. “My existence is based on the fact that there are schools and institutions that are not built with the children that I serve at the forefront. And so that’s what I’m working to improve.”

In yet another of his unending twists on trade, President Donald Trump has suggested he may not renew the current United States trade agreement with Canada and Mexico. Especially interesting was one of his explanations of why. “We don’t need anything that Canada has; we don’t need anything that Mexico has, but they need everything that we have, and they have to treat us better,” said Trump. “We don’t need their cars; we don’t need their lumber; we don’t need their energy; we don’t need anything that they have.” 

According to the numbers from April of this year, Canada is the leading trading partner of the US in imports and exports of goods, totaling $86 billion. Mexico ranks second, totaling $64.8 billion in two-way goods trade with the US. Overall, the regional economic integration of the United States–Mexico–Canada Agreement (formerly the North American Free Trade Agreement before Trump made a few changes in it and relabeled it as the USMCA in his first presidential term) has made all three parties to the agreement more competitive domestically and internationally. 

It is not remotely credible to suggest that the United States needs nothing from its two top trading partners any more than it would be credible to suggest that they need nothing from us. Trump’s dismissal of what Canada and Mexico have to offer Americans in trade is reminiscent of a famous communication sent to King George III by the Xianlong Emperor of the Qing Dynasty in China in 1793. The British Empire had sent a mission to Beijing, led by Lord George Macartney, to open China to British trade. But the Chinese emperor was not interested. 

In his edict delivered to the British king, he said, with some disdain, “I see no value on objects strange or ingenious, and have no use for your country’s manufactures .… It behooves you, O King, to respect my sentiments and to display even greater devotion and loyalty in future ….” (See Harry G. Gelber, The Dragon and the Foreign Devil: China and the World, 1100 B.C. to the Present (New York: Walker and Company, 2007, 164). 

The dismissal of the economic gains that result from trade was one reason why the Qing Dynasty was the last Chinese dynasty. What followed were two more centuries of Chinese economic stagnation and decline before a more trade-minded Chinese leader, Deng Xiao Ping, assumed power in the 1980s. 

Let’s hope that the same does not happen to the United States because President Trump has no interest in importing goods from the two countries that send us the most and thus contribute significantly to the prosperity of the American people.

The latest Long-Term Trend results from the National Assessment of Educational Progress (NAEP)—dubbed the Nation’s Report Card—are out, and they’re kind of “meh.” On one hand, 9‑year-olds posted modest gains in reading and math compared with the last assessment. But those gains only partially recover losses from the pandemic era, while 13-year-olds remain stuck well below where they were just a few years ago.

Perhaps most troubling, reading performance among older students remains where it was when the test was first administered in 1971, while math is only slightly higher than the first test in 1973. These results are particularly shocking when you consider that we spend three times as much per student as we did in 1970 (adjusted for inflation). According to the latest national figures, average per-student spending in public schools tops $20,000. 

Despite the Nation’s Report Card moniker, NAEP is not a report card for any particular school, teacher, or student. It provides a snapshot of how students are performing nationally. NAEP can tell us whether outcomes for various groups are improving or deteriorating, but it can’t tell us why.

The latest results show that decades of top-down reforms have not produced the improvements many predicted. Policymakers have repeatedly increased spending, expanded regulations, adjusted standards, redesigned assessments, and implemented various accountability systems. Yet long-term results remain disappointing.

That reality should make us cautious about claims that the next bureaucratic fix will finally solve the problem.

The deeper issue is accountability. In a monopoly system, leaders attempt to drive accountability by having government officials measure performance and impose consequences. But success depends on politicians and bureaucracies correctly identifying problems and acting on them. History suggests that is easier said than done.

Parental accountability works differently. Families don’t need years of NAEP data to know when their child is thriving or falling behind. They experience the results directly. When parents have meaningful educational options, schools have a stronger incentive to respond because dissatisfied families can leave—and take their funding with them.

The latest NAEP results don’t prove that school choice is the answer to every educational challenge. But they illustrate something parents already know—the existing system is often not generating the outcomes families deserve. If decades of centralized reforms have produced stagnant achievement and widening concerns about student performance, it makes little sense to double down on the same approach.

Ultimately, the latest NAEP results illustrate that millions of kids, particularly older ones, are stuck in a system that hasn’t meaningfully improved in half a century. Test scores are just a symptom. While not a panacea, educational freedom is a key part of the remedy.

If you’ve followed the Section 702 reauthorization debate this week, you’ve heard a parade of warnings about imminent intelligence darkness. Speaker Mike Johnson says the program cannot go dark. Defense Secretary Hegseth warns that allowing Section 702 to lapse will be “devastating to our national security.” Senator Grassley invokes the Taylor Swift Vienna concert plot. President Trump himself has tied reauthorization to safety at the World Cup and America’s 250th anniversary celebrations.

It’s a deliberately crafted scaremongering narrative. It also isn’t true. 

Here’s what actually happens if Section 702 lapses at midnight Friday, which seems inevitable now that the House this morning defeated another attempt to extend the program on a bipartisan vote of 198–218.

First, it is Title VII that sunsets, not just Section 702.

The sunset provision in the FISA Amendments Act applies to all of Title VII of FISA, not just Section 702. That distinction matters because Title VII contains more than the 702 program. It also contains authorities that the government uses to target Americans who are located abroad. When the clock runs out, 702 and the overseas-US-person authorities lapse together. The pending Senate bill reflects this: its extension section amends the Title VII sunset date, not a 702-specific one.

So a lapse is, if anything, broader than the “702” shorthand implies. What it does not touch is the rest of FISA.

The bulk of FISA is entirely unaffected.

Titles I and III of FISA—traditional electronic-surveillance and physical-search authority—are not part of Title VII and do not sunset. Neither do the pen-register, trap-and-trace, or business-records authorities nor the emergency provisions of the original 1978 Act. Any genuinely urgent collection against a specific target can proceed through an individualized court process under Title I, which is exactly what the Fourth Amendment contemplates and what the reform coalition has urged for years.

The “emergency” scenario the intelligence community invokes is one in which a target is important enough to justify days of apocalyptic floor speeches, but somehow not important enough to support a Title I probable-cause application. That contradiction deserves to be stated plainly.

Existing collection is grandfathered into 2027.

Section 702 operates under annual programmatic certifications approved by the Foreign Intelligence Surveillance Court (FISC), together with the directives served on providers under them. Under the FISA Amendments Act’s transition provision, acquisitions authorized by certifications and directives in effect at the moment of sunset may continue until those certifications expire. The FISC approved the current certifications in March 2026; they run until roughly March 2027.

Representative Jamie Raskin made the point on the House floor Wednesday: government surveillance activities will continue unchanged after Friday, because everything already authorized and certified is “already in motion” and continues, in his words, at least through March 17, 2027. A senior Republican committee aide told WIRED the same thing in blunter terms: the program has the FISA court’s permission to continue for another year, so it will continue whether Congress acts or not—and none of the members claiming the program ends Friday will be calling it dead on Monday.

Executive Order 12333.

The overwhelming bulk of overseas signals intelligence never depended on Section 702 in the first place. It runs under Executive Order 12333, the daily operating charter for the executive branch’s intelligence components, which requires no statute and no FISC order. A Title VII lapse removes not one 12333 collection platform.

This is not a theoretical point, and I have documented it before. A CIA Inspector General audit dated August 7, 2002, covering the years 1995 through 2000—a document the Cato Institute had to sue the Privacy and Civil Liberties Oversight Board under FOIA to obtain—shows the CIA apparently conducting precisely the kind of internet “backbone” surveillance now carried out under Section 702, with no judicial oversight whatsoever, years before Section 702 existed. The program captured and retained U.S.-person data, presuming targets were foreign until it discovered otherwise. If 702 lapses, that history is a reminder that the underlying collection capability could potentially be resumed under 12333. The capability predates the statute; it does not depend on it.

Human intelligence, open-source collection, and liaison relationships are untouched.

Human-source intelligence, open-source and social-media collection, and the Five Eyes and broader intelligence partner relationships all continue without interruption. Foreign partners do not stop warning the United States about threats to American lives because a domestic statute expires.

What a lapse actually costs—stated honestly.

There is one genuine operational seam, and reform advocates do themselves no favors by pretending otherwise. 

Section 702’s compelled-assistance mechanism—the authority to serve new directives on US electronic communication service providers, backed by statutory liability immunity for their cooperation—lapses with Title VII. The government could not issue new directives or add new certifications during a lapse. A genuinely new foreign target whose communications transit a US provider would require an existing directive covering that service, a 12333 access point, or an individualized FISA process; providers may decline new requests absent the statutory compulsion and immunity. Even the IC’s own witnesses concede the database itself remains available to search; the honest version of their complaint is that the data grows stale and new compelled assistance pauses, not that the country goes blind.

That sentence does not generate the urgency needed to push through a warrantless reauthorization, which is why no one in the administration is saying it.

The framing obscures a real danger—pointed the other way.

When Title VII lapses, the collection capability survives through 12333 and the surviving FISA titles, but the oversight architecture built specifically around 702—FISC programmatic review, minimization procedures, congressional reporting, and the compliance regime—goes with the statute. An intelligence community that keeps the capability while shedding the oversight is not a comforting outcome, but it is entirely the fault of Trump and his senior advisors, as well as House and Senate leadership.

The infrastructure that would let the public verify the government’s compliance claims has meanwhile been hollowed out. The president fired the three Democratic members of the Privacy and Civil Liberties Oversight Board in January 2025—a removal a federal court later held unlawful, in litigation still ongoing—leaving the board below quorum and its sole functioning member a Trump appointee. The FBI’s internal office responsible for surveillance oversight was reportedly disbanded by Director Kash Patel—the same official who built his career attacking the FBI’s abuse of Section 702 and whose agents, this spring, queried FBI databases for information on a New York Times reporter who wrote an article he disliked.

And the case for the reform at the center of that fight—a warrant requirement before the government searches Americans’ 702 data—has never been better supported by the public record. In an April 2 statement, Senator Ron Wyden noted that “sensitive” warrantless searches—those involving American journalists and political and religious organizations—more than tripled in 2025, with the FBI refusing to explain why. Wyden has separately said a still-secret March 2026 FISC opinion describes serious abuses; the administration has refused to declassify it even after the Senate Intelligence Committee’s bipartisan leadership jointly asked for its release.

Congress is being asked to extend, without a warrant requirement, a surveillance program whose compliance record cannot be independently verified, whose oversight bodies have been deliberately disabled, and whose custodians have shown a personal willingness to turn its tools on the press.

The program is not going dark on Friday. The real question is whether the House and Senate leadership will finally tell Trump that embracing the Fourth Amendment’s warrant requirement is the only way to end this saga.