PPI Names Astri Kimball Van Dyke Senior Fellow for Artificial Intelligence Policy

WASHINGTON (August 27, 2026) — Today, the Progressive Policy Institute (PPI) announced that it has named Astri Kimball Van Dyke a senior fellow for AI policy to serve in the Capitol Hill office of Rep. Don Beyer (D-Va.). In her role, Van Dyke will focus on how Congress can build the technical fluency and legislative capacity to aid and regulate the development of AI and related infrastructure in the United States.

“Congress will play a critical role in making sure AI benefits all Americans. Members of Congress don’t need to be experts when it comes to AI, but the American people deserve members and staff who are fluent in these tools and understand how the technology stack works,” said Van Dyke. “Congressman Beyer has been a leader on this issue, getting a degree in machine learning while representing his district. Leadership on AI policymaking is up for grabs, and state legislatures aren’t waiting for the federal government. Alongside other Congressional offices and my colleagues at PPI, I’m looking forward to working on policy that can keep pace with a rapidly-changing field, so Washington is equipped to lead and Americans everywhere can share in the benefits of this pioneering technology.”

Prior to being named a PPI senior fellow, Van Dyke spent 12 years at Google, most recently leading its global competition policy team, after serving as a political appointee in the Obama administration, including as deputy counsel to then-Vice President Joe Biden.

“Astri is exactly the kind of expert Congress needs right now. Her approach to public policy from her days in the Obama White House to today is precisely the experience Congress needs to develop smart AI policy,” said Lindsay Mark Lewis, CEO of PPI. “Members of Congress need to get AI policy right, to show Americans the real upside it can offer across health care, government services, education, and more.”

PPI is the first think tank to embed a fellow for AI policy expert directly in a congressional office. This new role complements PPI’s existing congressional policy fellowships, which place staff with the New Democrat Coalition and Blue Dog Coalition.

Founded in 1989, PPI is a catalyst for policy innovation and political reform based in Washington, D.C. Its mission is to create radically pragmatic ideas for moving America beyond ideological and partisan deadlock. Find an expert and learn more about PPI by visiting progressivepolicy.org. Follow us at @ppi.

###

Media Contact: Ian O’Keefe – iokeefe@ppionline.org

The Trump administration should not get new tariff authority

The Trump administration should not get new tariff authority

THE NUMBERS: U.S. tariff collection on hockey pads, helmets, sticks, pucks & pads –

2025 $40 million
2024 $8 million

WHAT THEY MEAN: 

Assessing the proposed “Lindsey Graham Sanctioning Russia and Ukraine Act” this week, PPI’s experts — Kyiv-based New Ukraine Project Director Tamar Jacoby, Vice President Ed Gresser, National Security Director Danielle Steitz — encourage its effort to strengthen financial sanctions on Russia and limit the ability of its creaky “shadow” fleet of tankers to ferry oil around the world. By contrast, the three pan its tariff measures. As passed by the Senate early this month, these offer the Trump administration a wholly new power to impose tariffs of up to 100% on goods from an uncertain array of countries involved in buying Russian oil or gas, or in some mostly undefined way helping Russian energy get around existing sanctions. (Some explanation below.) PPI comment:

“The bill’s tariff section is unnecessary — the administration already has sanctions power under the International Emergency Economic Powers Act — and dangerous. Unchecked authority to impose new tariffs on vague grounds, after the administration’s extensive and continuing misuse of laws such as Section 301 and Section 232, poses a high risk that Mr. Trump would use it not to pressure Russia but to further harass America’s neighbors and allies, and impose more costs on the American economy.”

The past week’s experience illustrates the point vividly. For quite unclear reasons, the Trump admin. chose last Thursday to detonate a northern-border crisis with Canada, using a never-tested, 97-year-old Hoover administration law (“Section 338”) to impose 50% tariffs on about $20 billion worth of Canadian goods. Having failed to extract concessions, these are now in effect; Canada’s retaliation, targeting an identical amount of American steel, aluminum, seafood, dairy, appliances, and so on, comes after Labor Day.

Setting aside the unquantifiable but likely high strategic/security/etc. cost of alienating a neighbor, and the moral cost of pointlessly picking a fight with a friend, the purely economic shrapnel will hit home. Higher U.S. tariffs mean Americans will pay higher prices for tariffed goods. Higher Canadian tariffs on American goods, since Canada is usually the top buyer of made-in-America manufactured goods and ranks second for farm products, means Canadians will pay more for American goods, likely buy less of them, and U.S. export industries will struggle. Case in point: Canadians buy half of Maine’s lobster catch each year, about 20,000 tons a year, and Tuesday’s Canadian retaliation means they now have to pay a 25% tariff on each crustacean.

Here’s a less obvious example: hockey parents and high school athletic departments in Michigan, Minnesota, Wisconsin, Maine, Vermont, Massachusetts, and other northern tier states should expect some price hikes in hockey gear, and prepare themselves for possible spikes in rink fees. A Hockey News discussion last week is a point of departure:

“New tariffs imposed by American President Donald Trump include hockey sticks and equipment. The result will be members of USA Hockey paying more to play the game, which could slow the sport’s record growth in the USA following double Olympic gold and the PWHL’s [Professional Women’s Hockey League] rise. The new 50% tariffs, which will take effect August 19, just before many hockey teams and leagues begin to hit the ice, will make ice hockey significantly more expensive for youth and families looking to get into the sport in the United States.”

How so? Fitting a teen out for the rink is pricey. Think $200 for skates, $70 for a college-quality carbon-fiber stick, $200 for pads, $100 for gloves, $30 for a bucket of pucks, more for goalies. Rink managers, meanwhile, need expensive machinery to smooth their ice (a Zamboni machine runs from $125,000 to $180,000), keep air circulating at the right temperature, and sharpen players’ skates before games. This means teams have to defray ice-time costs through fees. Maine’s middle-school program charges $345 per player plus $55 for association membership; in Alpena, Michigan,  on Thunder Bay — where median family income is $47,500 — youth hockey fees run from $500 for “mites” to $1,000 for teens.

Under the Congressionally authorized “Harmonized Tariff System”, hockey gear is usually free. Last year’s tariffs on Asian consumer goods (where North American hockey equipment firms like CCM and Bauer do most of their assembly) helped hike equipment prices by about 10%. Canadian gear, still exempted then under the “USMCA” duty-free rules, made up about a fifth of U.S. hockey imports last year. That was $25 million worth of pads, helmets, pucks, and sticks. So at face value, Thursday’s 50% tariff will cost American hockey parents and athletic departments about $12 million. If the crisis escalates to the high-priced products rinks need to run — Zambonis get ambiguous treatment under Mr. Trump’s new tariff order — rink operation costs will spike, ice-time prices will follow them up, and team membership fees will rise with them.

Now back to the Russia sanctions and tariff bill.

As Jacoby, Gresser, and Steitz argue, the proposed sanctions will help, and ought to go ahead. (See Jacoby’s deep-dive piece on sanctions this week in the Washington Monthly.) The bill’s tariff clauses, by contrast, are risky rather than useful. They authorize new tariffs of up to 100% on goods from (a) the top-five buyers of Russian oil, (b) the top-five buyers of Russian gas, and (c) a murkily defined top-five “facilitators of energy sanctions evasion.” Complex accounting for energy purchases, and even less precise definitions of “facilitation,” make this authority easy to redirect onto friends and allies. And as this authority would come as part of a newly passed law, rather than one of the 50-to-100-year-old statutes the administration has exhumed for its tariff decrees to date, any resulting tariffs would likely be legally stronger than those the administration has tried so far.

So: On the tariff matters, the main need is (as Sen. Ron Wyden, D-Ore., and Rep. Linda Sanchez, D-Calif., have proposed) to impose discipline and restore constitutionally appropriate policymaking, by requiring administrations hoping to impose tariffs to get Congressional votes of approval before they can act. For this particular bill, Congress should at minimum require the same for any use of tariff authority — that is, no tariff would go into effect until Congress had voted to approve it — or scrap the tariff piece altogether.

FURTHER READING

Russia sanctions/tariff bill (1):

Legislation text.

PPI’s Tamar Jacoby, Ed Gresser, and Danielle Steitz approve of new sanctions, pan new tariff authority, and call for speeding up military aid to Ukraine.

And from Kyiv, Jacoby’s in-depth assessment in the Washington Monthly this Monday.

A warning:

Trump admin.’s August tariff decree with 50% tariffs on $20 billion worth of Canadian-made autos and parts, liquor and beer, hockey equipment, machinery, auto parts, etc. For context, U.S. imports of Canadian goods last year totaled $382 billion. The resource side of the trade relationship — fertilizer and energy — is still mostly exempt, except for wood.

Stunned Hockey News staff and readers assess the impact.

Toronto-based BladeTech Hockey explains why high-performance sticks are so expensive. (TL/DR: “materials, manufacturing complexity, R&D amortization, and 2026 tariff impacts.”)

Canada’s retaliation targets American lobsters, air conditioners, cheese, hand tools, steel and cars.
In Lansing, Gov. Whitmer tallies damage-to-date on Michigan inflation and exports.

Last January’s Canadian Broadcasting Corp. report from a worried U.S.-Canada lobster conference.

And alarmed reactions from the Maine political world.

Sanctions/tariff bill (2):

The bill’s sanctions add Russian officials and enterprises to the lists of individuals and industrial sectors covered by financial sanctions; ban U.S. imports of uranium (but not fertilizer, where U.S. imports are running at $200 million a month this year); and ban investment in Russian energy. It isn’t automatic, though, as the bill also allows administrations to waive any of these sanctions.

The tariff piece ‘authorizes’ tariffs of up to 100% on the five countries topping the lists of buyers of Russian gas and Russian oil, and the five countries contributing most (in some vague and undefined way) to energy sanctions evasion. The top three energy buyers are China, India, and Turkey – a running tally from the Finland-based Center for Research on Energy and Clean Air estimates that these three together account for 70% of Russia’s roughly $250 billion in annual energy sales – but the fourth- and fifth-largest buyers aren’t as easy to identify, especially with separate lists for oil and gas. ‘Sanctions evasion’ is not clearly defined, and could be used to target all sorts of countries.

And recommended reading:

Poland’s Institute for International Affairs has some creative thinking on the “shadow fleet” and energy-sanctions policy.

ABOUT ED

Ed Gresser is Vice President and Director for Trade and Global Markets at PPI.

Ed returns to PPI after working for the think tank from 2001-2011. He most recently served as the Assistant U.S. Trade Representative for Trade Policy and Economics at the Office of the United States Trade Representative (USTR). In this position, he led USTR’s economic research unit from 2015-2021, and chaired the 21-agency Trade Policy Staff Committee.

Ed began his career on Capitol Hill before serving USTR as Policy Advisor to USTR Charlene Barshefsky from 1998 to 2001. He then led PPI’s Trade and Global Markets Project from 2001 to 2011. After PPI, he co-founded and directed the independent think tank ProgressiveEconomy until rejoining USTR in 2015. In 2013, the Washington International Trade Association presented him with its Lighthouse Award, awarded annually to an individual or group for significant contributions to trade policy.

Ed is the author of Freedom from Want: American Liberalism and the Global Economy (2007). He has published in a variety of journals and newspapers, and his research has been cited by leading academics and international organizations including the WTO, World Bank, and International Monetary Fund. He is a graduate of Stanford University and holds a Master’s Degree in International Affairs from Columbia Universities and a certificate from the Averell Harriman Institute for Advanced Study of the Soviet Union.

Read the full email and sign up for the Trade Fact of the Week.

PPI: Yes to Stronger Sanctions on Russia and Support for Ukraine; No to Giving the Trump Administration New Tariff Authority

WASHINGTON (August 25, 2026) — Today, Progressive Policy Institute (PPI) experts on Ukraine, trade policy and national security issued the following statement on the Lindsey Graham Sanctioning Russia and Iran Act, applauding the additional financial and energy-market pressure the bill may put on Russia, but criticizing its very broad tariff authority and arguing that it should either be dropped or sharply circumscribed.

PPI’s Kyiv-based New Ukraine Project Director Tamar Jacoby, who offered extensive assessment of the bill in the Washington Monthly, said:

“Nearly five years of war — including a summer of spectacular Ukrainian strikes on Russian oil refineries — have done serious damage to the Russian economy. The US can and should be doing more to help raise the cost for the Kremlin, and the Graham bill makes an important start. Now the House has an opportunity to improve the package by clarifying what Trump can and cannot do to implement it — strengthening the sanctions provisions while limiting the all-but unchecked tariff authority it hands the president.”

PPI Vice President and Director for Trade and Global Markets Ed Gresser said:

“With Russia’s military exhausted from its four-year full-scale invasion of Ukraine, Congress is right to increase economic pressure on Russia and countries supporting it. The bill’s financial and other sanctions are good steps. We can and should go further by banning Russian fertilizer — the U.S.’s major remaining import from Russia, now running at an average of $200 million a month this year. By contrast, the bill’s tariff section is unnecessary — the administration already has sanctions power under the International Emergency Economic Powers Act — and dangerous. Unchecked authority to impose new tariffs on vague grounds, after the administration’s extensive and continuing misuse of laws such as Section 301 and Section 232, poses a high risk that Mr. Trump would use it not to pressure Russia but to further harass America’s neighbors and allies, and impose more costs on the American economy.  Last Thursday’s fiasco with Canada — a totally unnecessary breach with a friendly neighbor, with tariffs likely to cost Americans billions of dollars this fall — is a case in point. If this final bill is to include any tariff authority at all, therefore, the House should at minimum revise it to require a final Congressional vote of approval for any tariff increase on any country.”

PPI Director of National Security Policy Danielle Steitz said:

“The brave citizens of Ukraine have spent more than four years defending their nation from Russia’s illegal invasion, and it is heartening to see the U.S. Congress continue to send clear bipartisan signals of support. However, it is undeniable that President Trump has demonstrated his disregard of congressional intent time and time again, particularly around the topic of support to Ukraine. Just last month it was reported that the Department of Defense will not finish spending the $400 million that Congress authorized for Ukraine military assistance until fiscal year 2029, an unacceptable slow-rolling of aid that has rightly sparked bipartisan criticism on Capitol Hill. Instead of hoping that this time will be different, the House should adopt common-sense guardrails asserting Congress’s tariff powers and making its intent crystal clear: that this legislation’s purpose is to impose costs on Russia and to support Ukraine.”

Background: The Senate passed S. 5025, the Lindsey Graham Sanctioning Russia and Iran Act, on August 6, 2026, and the House of Representatives is preparing to consider it after its return from the August recess. The bill imposes new sanctions on Russian officials, businesses, and financial service providers, bans imports of Russian uranium (though not fertilizer, the largest single U.S. import from Russia). It also provides the Trump administration authority to impose tariffs of up to 100% on goods from three sets of countries: the five largest buyers of Russian oil, the five largest buyers of Russian natural gas, and the five countries most involved in helping Russia in some unspecified way avoid sanctions. With the House vote likely in early fall, PPI experts and Ukraine-watchers argue that while strengthening sanctions on Russia is urgent, the bill’s nearly unlimited tariff authority is unnecessary and invites further abuse by the Trump administration. They urge the House to make significant changes to the tariff section, and tighten U.S. sanctions further, before passage.

Founded in 1989, PPI is a catalyst for policy innovation and political reform based in Washington, D.C. Its mission is to create radically pragmatic ideas for moving America beyond ideological and partisan deadlock. Find an expert and learn more about PPI by visiting progressivepolicy.org. Follow us at @ppi.

###

Media Contact: Ian O’Keefe – iokeefe@ppionline.org

Manno for CC Daily: The future of learning is moving into the workplace

Community colleges aim to connect the education they offer to work. Now the workplace itself is a major site of learning. That creates an opportunity and a challenge for community colleges.

Artificial intelligence, workforce platforms, professional software and employer training systems increasingly help people acquire skills while performing real tasks. These tools teach workers to use new technologies, manage projects, solve problems and assume greater responsibility.

But learning often remains invisible outside the company. It may not translate into college credit, a recognized credential or a clear route to a better job. That’s why a new European investment report has insights that should matter to American community colleges.

Read more in CC Daily

Manno for Real Clear Education: As Students Return, What Did Federal Pandemic Aid Accomplish?

As students return to classrooms, the educational effects of the pandemic remain unfinished business. More than six years after schools first closed, achievement remains below 2019 levels in many districts. Researchers are still trying to determine how much the nearly $190 billion in federal emergency aid helped students recover.

A new working paper from the National Bureau of Economic Research (NBER) appears to offer a discouraging verdict. It finds that additional federal aid did not increase school spending or improve student test scores in the districts the researchers studied.

That conclusion seems to conflict with research from the Education Scorecard and the Center for Analysis of Longitudinal Data in Education Research, or CALDER. Both found that federal pandemic aid contributed to academic recovery, particularly in mathematics and in high-poverty districts.

But the studies are not as contradictory as they first appear. Together, they offer a more useful lesson. It’s not enough to ask whether money matters. Policymakers must ask whether federal aid becomes additional education spending, where it goes, and whether schools use it for activities likely to improve learning.

Read more in Real Clear Education

Gresser in The Wall Street Journal: How Will the New Canadian Tariffs Affect the U.S. Economy?

[…]

Ed Gresser, a former assistant U.S. trade representative now working at the Progressive Policy Institute, said that was clear from the items Canada chose, including lobster and seafood. Half of the lobster caught in Maine goes to Canada, he said, and Maine has an important election this fall that will help determine control of the Senate.

“The states that are most reliant on Canada as an export market are often the northern-tier states—Maine, Michigan, Minnesota, Wisconsin, New Hampshire,” Gresser said. Canada is “trying to show the Republican party that there’s a systemic cost to doing this sort of thing.”

[…]

Read more in The Wall Street Journal

Moss for Competition Policy International: “Anti-Monopoly” Antitrust Enforcement: Lessons Learned from the Biden Administration

In this paper, Diana L. Moss reflects on the legacy of the Biden Administration’s “Anti-Monopoly” agenda and its implications for the future of U.S. antitrust enforcement. While recognizing the movement’s ambition to challenge concentrated economic power, she argues that its departure from the consumer welfare framework and its reliance on expansive legal theories ultimately produced limited success in the courts. The article offers a broader assessment of how ideological swings in enforcement philosophy can affect both legal certainty and the long-term credibility of antitrust institutions.

I. THE ANTI-MONOPOLY MOVEMENT IN CONTEXT

In 2021, the Biden administration installed “Anti-Monopoly” advocates at the U.S. Department of Justice (“DOJ”) and Federal Trade Commission (“FTC”). The movement, which emerged in the mid-2010s, is rooted in the idea that antitrust is a critical tool for solving the economic, political, and social problems, and threats to liberty, that flow from corporate power. Anti-Monopoly ideology rejects the consumer welfare standard that has played a central role in antitrust for decades, advocating instead for bright line tests for corporate “bigness.”

The practical effect of the Anti-Monopoly movement’s repudiation of the consumer welfare standard is the dissolution of antitrust’s “effects-based” approach for determining violations. That analysis links anticompetitive consolidation and business practices to harm to consumers, through higher prices, lower quality, less variety, or slower innovation. Without effects-based analysis, decisions on antitrust violations would revert simply to thumbs-up or thumbs-down findings based on mandated threshold criteria.

The role and interpretation of the consumer welfare standard has been vigorously debated for decades. The standard earned a questionable reputation under conservative “Chicago School” doctrine that commandeered the antitrust enterprise for almost 40 years beginning in the 1970s. For example, laissez faire enforcers erroneously focused almost exclusively on the short-term price effects of anticompetitive consolidation and conduct. Such threats are easily overcome by the promise of efficiencies, such as cost savings or expanded R&D capability, from mergers or business practices that otherwise limit competition. The damage to competition and consumers wrought by Chicago-School conservatism is still felt by consumers today.

Backlash to lax Chicago-School enforcement sparked the center-left “Post-Chicago” movement in the late 1990s. The Obama administration issued the first ever Executive Order on Competition,3 pursued aggressive cartel enforcement, and withdrew conservative past policy guidance on monopolistic conduct. On the merger front, Obama administration antitrust enforcers blocked highly concentrative mergers such as AT&T-T-Mobile (2011), Staples-Office Depot (2016), Sysco-U.S. Foods (2015), Halliburton-Baker Hughes (2016), Aetna-Humana (2016), and Anthem-Cigna (2017).  These deals eliminated head-to-head competition or encouraged anticompetitive coordination, invoking the “structural presumption” of illegality, under which no efficiencies claims could possibly save the day.

Against the backdrop of major shifts to far-right antitrust conservatism and far-left antitrust progressivism over the last 50 years, the pragmatism and effectiveness of the center-left Post-Chicago approach is under-recognized. For Anti-Monopoly advocates, deployment of the full scope of the consumer welfare standard and the structural presumption (to knee-cap efficiencies defenses) was not really relevant. Rather, embracing the Chicago School’s inaccurate interpretation of the standard and deference to efficiencies is central to the Anti-Monopoly argument that the standard should be rejected outright.

Control of the enforcement agenda under the Biden administration thus provided a platform to advance wholesale antitrust reform. This article empirically unpacks results from the Biden enforcement era, strengthening the case for why swings in antitrust ideology impose risks and high opportunity costs on enforcement, with a direct impact on promoting competition and protecting consumers.

Read the entire publication

Jacoby in CEPA: Ukraine’s Defense Industry: The Motor of Post-War Revival

[…]

The shape of a defense-led recovery depends on whether Ukraine makes the right deals in Europe as it seeks to maximize the value of its expertise, according to Tamar Jacoby, director of the Progressive Policy Institute’s New Ukraine Project.

The defense sector has become increasingly decentralized to allow it to be flexible and innovate with real-time feedback from the frontline, but it means it often misses out on economies of scale.

Ukraine’s Ministry of Defense has also, until recently, banned weapon exports, and deals that have been permitted so far have been time-consuming and stymied by bureaucracy. The businesses that flourish tend to be the ones with the right connections, Jacoby said.

Regulations need to be reformed to speed up processes, tackle corruption, and encourage the development of a robust defense export industry, which will attract talent.

Jacoby compares Ukraine to Israel, which was able to build on its military experience to fuel an economic and technological boom through sales of high-end, cutting-edge weaponry, supported by a government apparatus actively pushing exports.

[…]

Jacoby for Washington Monthly: Beyond Virtue Signaling on Ukraine and Russia

Just three days after the U.S. Senate voted overwhelmingly to advance the late Senator Lindsey Graham’s sweeping sanctions bill designed to force Russia to end the war in Ukraine, Vice President JD Vance called Ukrainian President Zelensky and asked him to stop bombarding oil tankers and terminals in the Russian Black Sea port of Novorossiysk.

A key backer of the Graham bill, Republican Katie Britt of Alabama, defended the statute passionately on the Senate floor. “We’re going to cut off the flow of money to [Russian dictator Vladimir] Putin,” Britt declared emphatically. But the international partnership that owns the tankers and terminals Vance asked Zelensky to spare, the Caspian Pipeline Consortium (CPC), generates more than $1.4 billion a year for, among others, the Russian Federation, which owns or controls nearly a 50-percent share.

It’s not technically Russian oil that passes through the CPC’s thousand-mile pipeline and sprawling terminal complex—it’s crude from the vast oilfields in western Kazakhstan. But Russia profits handsomely from dividends, transit tariffs, and port fees. What’s complicated—and the reason Vance picked up the phone—is that U.S. energy giants Chevron and ExxonMobil also own significant shares of the Caspian pipeline.

Read more in Washington Monthly

The Next Phase of the AI Revolution

AI models have clearly transformed the process of writing software. Nevertheless, jobs in the U.S. software industry have risen over the past year, according to the Bureau of Labor Statistics, while the 12-month moving average of workers in computer and mathematical occupations hit an all-time high in July 2026. Looking at a broader measure, the unemployment rate for young people (aged 20 to 24 years old) in July 2026 was only 7.1%, down from an average of 8.3% in 2025.  

But while fears of an AI-driven jobs apocalypse have dissipated for the moment, new concerns have arisen. As open-weight models — many from China — have become more widely available and narrowed the gap with closed-weight frontier models, people are worried that American individuals and firms will become more vulnerable to dangerous cyberattacks.  

At the same time, hundreds of companies and organizations, including Amazon, Google, and OpenAI,  have signed the Open Weights and American AI Leadership letter, arguing in favor of open-weight models as an important part of a strong AI ecosystem. The letter called for “expanding access to AI, encouraging competition, robust application layers, and giving Americans greater control over the technology they rely on.”  

In this context, enterprises and individuals are increasingly worried about  “AI sovereignty,” a broad term that addresses fears of being excessively dependent on third-party AI models. According to the Foundation for American Innovation, 56% of business owners and C-suite executives are “concerned that their competitors will benefit from AI models trained on their organization’s proprietary data, workflows, or institutional knowledge.” Nearly all agree that transparency is important, and AI providers should disclose how customer data is used to improve AI systems.  

AI sovereignty is also relevant for the ability of AI to tackle difficult problems in the physical world, including supporting national security, improving manufacturing and construction productivity, and lowering health care costs.  Getting a positive return on investment in these areas requires the accumulation and protection of enormous new granular and proprietary data sets, which includes the slow and expensive collection of data on rare but important long-tail events. 

The importance of AI sovereignty cannot be overstated, even though it’s tough to define exactly what it is. At one end of the spectrum, AI sovereignty can be defined as control over data, models, and infrastructure. For example, open-weight models running on self-owned hardware mean that there’s no possibility of leakage of private data and insights. At the other end of the spectrum, concerns about AI sovereignty are directly addressed by products such as Google’s Sovereign Cloud, Amazon’s AWS “digital sovereignty,” and Apple’s newly revamped Siri, which focuses heavily on privacy protection. Companies such as Palantir and Nvidia are building a “middle-layer” to allow enterprises to connect to AI models of their choice without exposing proprietary data. 

Training on the broad corpus of the literate Web was necessary to bootstrap the first stage of the AI revolution. But that was the low-hanging fruit. Outside of coding, most executives are dissatisfied with their return on AI spend and the lack of business value delivered in real-world use cases and workflows.  Even in the software domain, there’s a growing sense that enterprises are not yet getting enough bang for the buck. 

The next stage—solving difficult real-world challenges — requires training on data that is typically located in different parts of the business and in different formats. Businesses need some application layer to make sense of that existing data, and put it into common language.

Moreover, it’s turning out in many areas that there’s a lot of data that simply doesn’t exist yet—physical data, material data, manufacturing data, clinical data.  And giving enterprises the right incentives to spend heavily to collect that essential data will require some form of AI sovereignty to protect the data’s value. 

Let’s be clear here. We do not foresee the popping of an AI bubble. Nor are we calling for heavy-handed regulation or government control. But we must realize the AI revolution is about to move into the next phase.

Manno for Washington Monthly: Workforce Pell’s Missing Rudder

For decades, Washington has been generous to Americans who could spend semesters or years earning a college credential. It has done far less for those who couldn’t follow that path because they were raising children, working full time, or simply needed income sooner than a standard college degree or certificate could provide.

When those students managed to get any postsecondary credential, it was usually because they paid for it themselves, borrowed the money, or patched together support from an employer or a state program that Washington had no hand in. They were relegated to the least-developed corners of higher education, mostly workforce programs offered by community colleges or for-profit providers, where students were sometimes badly served and left with debt and a credential that never paid off.

They also rarely received the advising or other support available to degree-seekers down the hall. Often, the courses they completed did not even count toward a further credential.

That changed, at least on paper, in July 2026. Workforce Pell was created under the One Big Beautiful Bill Act. It made students in short-term job-training programs eligible for Pell Grants for the first time. The program stretches a financial-aid system built around fifteen-week college semesters to cover programs as short as eight weeks.

Read more in Washington Monthly

Internet traffic to reach 10 sextillion bytes next year

FACT: Internet traffic to reach 10 sextillion bytes next year.

THE NUMBERS: Information content of seven communications forms –

 

One-paragraph plain-text email  2,000 bytes
One-paragraph email with hypertext  30,000 bytes
Voice call 200,000 bytes
Photograph 2 million bytes
TikTok video 8 million bytes
Movie 1 billion bytes

WHAT THEY MEAN: 

How much information moves around the world? And how does it travel?

Vocabulary first: Internet-watchers count information flow in “bytes,” a term invented by IBM’s Werner Buchholz in 1961. One byte is the amount of information necessary to portray a single character on a computer screen. Each byte in turn is made of eight “bits”, referring to the “binary digits” — 1 or 0 — used in computer code. The eight-bit rule isn’t required by math or physics; Buchholz and his associates just thought the 256 possible eight-bit strings, from 00000000 to 11111111, would be enough to represent all the necessary letters, numerals, mathematical notations, currency signs, and other keyboard characters.

Following Buchholz’s lead, experts at the International Bureau of Weights and Measures have fended off an uncontrolled proliferation of zeros and commas by inventing new words every decade or so. “Petabyte” and “exabyte,” referring to quadrillions and quintillions of bytes, date to 1982. “Zettabyte,” meaning a sextillion bytes, debuted in 1991 — three years before the World Wide Web went live — and is now the one-digit way to record Internet data traffic.

Data traffic: The International Telecommunications Union’s estimates of annual data flow — all the information moving around the world, across borders and within countries — from last year to the launch of the Internet:

2025      7.9400000 zettabytes
2020      3.7000000 zettabytes
2015      0.8700000 zettabytes
2010      0.2400000 zettabytes
2005      0.0300000 zettabytes
2000      0.0010000 zettabytes
1990      0.0000002 zettabytes

The jump from 0.000002 zettabytes in 1990 to 7.94 zettabytes in 2025 represents a 40 million-fold increase in information flow — voice calls, videos, Zoom sessions, movie downloads, gaming, texting, etc. — from the first website posting in 1989 to the present, mirrored by a fall in the cost of moving it around.

How did this happen? Mainly by stringing a few hundred wires under the oceans.

Cables: The commonly used stat — submarine fiber-optic cables carry about 95%, or possibly 99% — of all data traffic seems slightly squishy. NOAA is confident about it, but Telegeography traces its last verification to a 2015 FCC report, and at least 10,000 satellites have launched since then. But experts are at least confident that while the decade’s satellite deployment filled in lots of Internet dead spots — deserts, remote rural areas, ships at sea, planes in the air — cables still carry the vast majority of the bytes. Here’s their evolution:

1956: The first transatlantic telephone cable — a copper wire named “TAT-1” — went live two months after Buchholz’s “byte” coinage. (Earlier cables dated back to 1858, but transmitted telegraph and Morse code only.) In modern terms, TAT-1 could handle 2.3 megabits of data per second, or about 300,000 bytes. In human terms, that meant a maximum of 36 simultaneous voice calls. The cable owners, a U.S./British/Canadian consortium, charged $12 for a three-minute call. Median income that year was $2,432, or $6.70 per day, so a short U.S.-U.K. call would have cost a typical person nearly two days’ pay.

1988: A generation later, the first fiber-optic submarine cable (TAT-8) had a capacity of 560 megabits per second — 100 times its copper ancestor — and could manage 40,000 simultaneous voice calls. This one had 29 joint U.S., British, Canadian, and European owners. They charged $1.00 per minute at off-peak times, and $2 at peak.

2000: At the millennium, TeleGeography’s second annual Submarine Cable Map recorded 100 active fiber-optic cables, with the four-cornered U.S.-to-Japan “Pacific Crossing,” able to move 640 gigabits of data per second, the top performer. Though voice calls were no longer the glamor product, Pacific Crossing could theoretically have managed 10 million voice calls. Transatlantic calls cost about 10 cents per minute.

2010: Unity Submarine, also a Japan-to-U.S. cable, had a 7.7 terabyte capacity — another tenfold jump, equivalent to 175 million phone conversations. In that glum financial-crisis year, Telegraphy counted 406 operating fiber-optic cables. The FCC gave up on tracking international phone costs, since so many were free with Internet service.

2026: There are now 694 cables. Anjana, a Spain-to-South Carolina wire supposed to go live this fall, can carry 480 terabits of information per second. This is equivalent to 7.5 billion simultaneous conversations, easily enough to accommodate all the people in the world, if all they wanted to talk at the same time. Average Internet access cost in the U.S. (allowing, more or less, all the calls you want, no matter how long) is about $80 per month or $2.65 per day. For someone earning the U.S.’ median income of $65,050 per year, or $178 per day, that would be about seven minutes’ worth of income.

Two thoughts:

Physical: A 10,000-kilometer trans-Pacific giant with a top-of-the-line 24 fiber pairs typically is about an inch in diameter, weighs about 100,000 tons, and costs a bit above $500 million. A short one, for example one crossing the English Channel to connect London and Paris, would weigh about 80 tons and cost around $50 million.

Content: As the byte-content of emails, voice calls, and videos suggests, the information flowing over cables and down satellite beams is mainly video. Per Satellite Today, YouTube videos take up 13% of fiber-optic cable traffic, Netflix shows 10%, Facebook messaging 6%, TikTok videos 4%, and everything else 67%.

FURTHER READING

Cables & data flow:

Telegeography’s Submarine Cable Map.

And the ITU’s estimates of Internet data traffic.

Policy:

Trade: The Bureau of Economic Analysis reports that in 2025, American exports of information and communications services, plus “digitally deliverable services” such as entertainment, news, telemedicine, architectural plans, and so on, came to $959 billion — a bit less than half of the $2.2 trillion in goods exports, and 3% of U.S. GDP.

Policy and its growth impact: Last year’s OECD/WTO review of digital data flows, their growth effects, and the impact of regulation concludes that the best approach (at least from a GDP growth point of view) is “free flow of data combined with clear regulatory policies.” A hypothetical zero-regulation approach — no privacy rules, no content moderation, etc. — forfeits trust and leads to a loss of 0.9% of world GDP, while an open Internet with generally agreed-upon regulatory principles adds 1.8%. “Geoeconomic fragmentation,” like zero-reg, costs 0.9% of world GDP; “data autarky,” an extreme version of geoeconomics, yields a financial-crisis-style loss of 4.5%.

People:

Also from the International Telecommunications Union, a count of Internet users over time:

2026      6.1 billion, three-quarters of humanity
2020      4.7 billion people
2015      3 billion people, as Chinese access surges
2010      2 billion people
2005      1 billion, with rich-country access above 50% and middle-income rising fast
2000      0.36 billion, mostly in rich countries
1995     0.05 billion people, most in the United States

Words:

The International Bureau of Weights and Measures is responsible for order-of-magnitude prefixes, and has already prepared for the next round. After “zettabytes” come “yottabytes.” The next two, coined in 2022 and available whenever needed, are “ronnabyte” for 1 octillion bytes and “quettabyte” for a nonillion bytes. Both date to 2022. The prefixes apply universally — tons, meters, liters, etc. The mass of the Earth, for example, is 6 zettatons, and that of the solar system two ronnatons. The Milky Way, whose mass is about a trillion times the solar system’s, will have to wait for three more prefixes.

The Poughkeepsie Journal’s 2019 obit of the unassuming Mr. Buchholz — a Holocaust survivor, pioneer of transistor-based computing, and “byte” inventor.

ABOUT ED

Ed Gresser is Vice President and Director for Trade and Global Markets at PPI.

Ed returns to PPI after working for the think tank from 2001-2011. He most recently served as the Assistant U.S. Trade Representative for Trade Policy and Economics at the Office of the United States Trade Representative (USTR). In this position, he led USTR’s economic research unit from 2015-2021, and chaired the 21-agency Trade Policy Staff Committee.

Ed began his career on Capitol Hill before serving USTR as Policy Advisor to USTR Charlene Barshefsky from 1998 to 2001. He then led PPI’s Trade and Global Markets Project from 2001 to 2011. After PPI, he co-founded and directed the independent think tank ProgressiveEconomy until rejoining USTR in 2015. In 2013, the Washington International Trade Association presented him with its Lighthouse Award, awarded annually to an individual or group for significant contributions to trade policy.

Ed is the author of Freedom from Want: American Liberalism and the Global Economy (2007). He has published in a variety of journals and newspapers, and his research has been cited by leading academics and international organizations including the WTO, World Bank, and International Monetary Fund. He is a graduate of Stanford University and holds a Master’s Degree in International Affairs from Columbia Universities and a certificate from the Averell Harriman Institute for Advanced Study of the Soviet Union.

Read the full email and sign up for the Trade Fact of the Week.

Jacoby for Washington Monthly: The Next Big Thing in the Ukraine War: Ground Drones

We arrived at the headquarters of Ukraine’s 3rd Assault Brigade’s NC-13 strike unit just after 11 p.m. The soldier escorting me tells me to turn off my phone to avoid giving away our location, and we walk in silence through the night, then enter a darkened building. Deep inside, the operations room is buzzing with activity. Double-decker computer screens line the black walls. Radios crackle, a commander is barking orders, and a knot of plainclothes drone operators huddles quietly in a corner, poring over daytime satellite images of nearby fields and forests. “They’re plotting tonight’s route,” my escort tells me. Three kamikaze ground drones—unmanned vehicles, each about as big as a medium-sized dog—will travel overland for five or six hours to blow up a Russian bunker.

The Ukrainian defense industry is so innovative and dynamic—evolving in just five years from a few old-fashioned, Soviet-era manufacturers to global vanguard and pacesetter—that it is difficult to track all the groundbreaking developments. Already this year, Kyiv has upended the war with a new generation of long-range drones and missiles that have knocked out over one-third of Russia’s oil refining capacity. Closer to the battlefield, new mid-range drones are disrupting enemy supply lines and support logistics, dramatically slowing Russia’s advance. Most impactful for the U.S. and its Middle East allies, Ukrainian drone interceptors—whose widespread use began just over a year ago—helped Gulf states survive the Iranian Shahed strikes that threatened their cities this spring.

Yet many strategists in Ukraine and the West are focused on a different cutting-edge development, with several predicting that 2026 will be “the year of the unmanned ground vehicle.”

Keep reading in Washington Monthly.

Canter and Manno for The 74: 6 Keys to Moving CTE From Pilot to Permanent Part of a Redesigned HS Experience

Media stories about the skills mismatch between applicant qualifications and employer expectations have been a regular feature for decades. Among the most recent entries are a Washington Post story titled “Why Recruiters Can’t Find Workers and New Grads Can’t Find Jobs,” which warned of the oncoming “largest labor shortage in history,” and a Forbes report on a survey from the National Federation of Independent Businesses that showed a majority of respondents seeking to hire in June couldn’t find qualified staff. One in five said this is their biggest issue as a small business owner.

Solving this predicament is one rationale for the strong political support and attention shown to work-based learning and high school career and technical education programs. But leaders running these programs report a quandary of their own: They can’t find enough employers willing to host interested high school students in meaningful roles, or sometimes enough students to fill available spots.

The usual explanation, that one side or the other isn’t trying hard enough, misses what’s really going on. The problem isn’t effort. It’s infrastructure.

Read more in The 74

An Update on EU App Economy Jobs

The rapid rise of AI has created great uncertainty about the future of the job market globally, ranging from forecasts of a sharp plunge in white-collar employment to predictions of job gains. Taking the middle ground, one recent report showed no evidence that AI adoption in the European Union was reducing employment in the short run. 

In this blog, we focus on one labor market topic in particular: The growth of App Economy jobs in the EU. PPI has been tracking EU App Economy jobs since 2016.  Our last published comprehensive estimates of EU App Economy jobs were released in November 2023, a year after the release of ChatGPT.  To summarize our results, PPI finds an estimated 2.94 million App Economy jobs in the EU as of 2025, up from a revised 2.75 million jobs in 2023 and 1.91 million in 2019.

The AI revolution has triggered a massive surge in mobile app submissions to the Apple App Store and Google Play. According to Appfigures, the number of new app submissions to the App Store rose by 24% in 2025. The trend has accelerated into 2026. In the first half of 2026, new apps doubled compared to a year earlier, according to the New York Times

The wave of new app creation is being driven in part by AI coding tools, which make it easier than ever for anyone to become an app developer. Equally important, more and more apps are incorporating AI as part of their essential functionality, which requires app developers to exercise their creativity and knowledge. 

For example, earlier in 2026, Vivino, the world’s leading wine app, was looking for an iOS engineer in Copenhagen to help build AI-powered product features into the app. Zing Coach, an AI-powered fitness application, was looking for an Android Engineer in Poland. Climate tech startup autarc was looking for a senior iOS engineer in Berlin to join their AI team, building apps that help energy installers speed up the process of designing and deploying customized heat pump and photovoltaic systems. 

Based on our analysis of job postings, combined with Eurostat employment data for “information and communications technology professionals,” PPI finds an estimated 2.94 million App Economy jobs in the EU as of 2025, up from a revised 2.75 million jobs in 2023 and 1.91 million in 2019 (table).  For this analysis, a worker is in the App Economy if he or she is in:

  • An IT-related job that uses App Economy skills — the ability to develop, maintain, or support mobile applications. We will call this a “core” app economy job. Core app economy jobs include app developers; software engineers whose work requires knowledge of mobile applications; security engineers who help keep mobile apps safe from being hacked; and help desk workers who support the use of mobile apps.
  • A non-IT job (such as sales, marketing, finance, human resources, or administrative staff) that supports core app economy jobs in the same enterprise. We will call this an “indirect” app economy job.
  • A job in the local economy that is supported either by the goods and services purchased by the enterprise, or by the income flowing to core and indirect app economy workers. These “spillover” jobs include local professional services such as bank tellers, law offices, and building managers; telecom, electric, and cable installers and maintainers; education, recreation, lodging, and restaurant jobs; and all the other necessary services.

To estimate the number of core App Economy jobs in the European Union, we combine multiple

sources of information in a systematic process, including an estimate of the share of tech job postings that use App Economy skills. We use the various country-specific job databases found at Indeed.com/worldwide. Indeed, which bills itself as “the No. 1 job site in the world,” offers a searchable, continually updated database of job postings for more than 60 countries. 

We also use Eurostat’s estimates of the number of information and communications technology (ICT) professionals in EU countries, which rose by 2.7% in 2025.  Then we use a conservative multiplier of indirect and spillover jobs to estimate overall App Economy jobs. A more detailed description of the original methodology is found in “The App Economy in Europe: Leading Countries and Cities, 2017” (October 2017) and “The App Economy in India” (September 2019). 

Our methodology also allows us to estimate the number of App Economy jobs associated with different mobile operating systems. We find that in 2025 the EU had 2.0 million jobs associated with the iOS ecosystem, and 2.3 million jobs associated with the Android ecosystem. (The two add up to more than the total number of App Economy jobs because many jobs belong to both ecosystems.)

 

 

 

 

Data: Eurostat, Indeed, PPI estimates and calculations
*Revised from previously published figures to account for new data. 

Malec for 535 News: Advanced Recycling Can Save American Families Millions, Congress Should Help Scale It

Every week, millions of Americans do their part to recycle: they rinse their plastic containers, sort them, and put them in recycling bins. It’s a simple routine built on the belief that small actions make a difference. But this good-faith effort from the public can only get us so far.

Today, less than 10 percent of plastic waste gets recycled. The rest goes to landfills.

The problem isn’t effort. It’s the system those efforts depend on. Traditional recycling systems were never designed to handle many of the complex and mixed plastic products we use today, including shopping bags, bottle caps, and food containers. As a result, most plastic waste still ends up in landfills, and every ton of it comes with hefty costs on everyday Americans.

Read more in 535 News