The World Cup Ends Today. The Surveillance Will Continue.

Good morning!

The World Cup final will get underway in just a few hours, and here in New York the excitement is palpable. It’s been a fantastic tournament—despite all the politics—and yesterday’s third place contest between England and France was a thriller. For those of you that are fans of the sport, I hope the final is as exciting.

But while the World Cup ends today, the increased surveillance won't, write Ranking Digital Rights’ Tayrine dos Santos Dias and Leandro Ucciferri. Their investigation, titled ”No Fair Play: Mapping the 2026 World Cup Surveillance Stack,” details the companies behind the data harvesting, predictive policing, and aerial surveillance systems installed across 16 host cities.

“We found that at least 21 companies helped build or enhance the surveillance structures used to monitor, track, and profile fans, visitors, and locals,” the authors write. “These vendors provide products and services that we classify into a five-layer stack: data harvesting at scale, predictive policing, communications interception, data integration and command centers, and aerial surveillance.”

You can read about it on Tech Policy Press here.

The United States Department of Homeland Security (DHS) pumped hundreds of millions into FIFA-related security, a point that is discussed in this week’s podcast. But these investments are just a fraction of the billions more in tech investment that is reshaping DHS and its component agencies into an engine of authoritarianism. Tech Policy Press contributing editor Dean Jackson talks with Surveillance Resistance Lab’s Mizue Aizeki & Just Futures Law’s Paromita Shah about the oligarchs behind ICE—and what it will take to reverse it. With Mijente, their organizations published a report last week titled “No Tech for Ice: Oligarchs, Immigration Enforcement, and the Threat to Democracy.”

Listen to the podcast here.

And while the World Cup final action will be broadcast to screens around the world today, much of the labor that underpins the sport will remain invisible. “Soccer’s datafication is creating a new industry—one that follows the model of the global AI industry: lucrative, concentrated, and unequal,” writes Tech Policy Press fellow Tatiana Dias. This World Cup's AI, sensors and VAR are celebrated as making the game smarter—but behind them sits a hidden labor chain.

Tatiana traces who does the work.

There’s a great deal more on Tech Policy Press this week.

Unanswered questions post-Geneva

Following the UN Global Dialogue on AI and related events that took place around it, a number of Tech Policy Press contributors wrote last week to share their analyses and perspectives on developments in Geneva.

• Geneva was a mirror, writes Konstantinos Komaitis, senior resident fellow at the Democracy and Tech Initiative at the Atlantic Council: a world that knows AI can't be governed alone but "lacks the courage to decide what governing together is actually meant to achieve." Until AI has a North Star, every UN dialogue just manages disagreement.
• The UN Independent International Scientific Panel on AI's assessment is a reference point for AI governance worldwide, writes Eryk Salvaggio, a Gates Scholar researching AI and the humanities at the University of Cambridge. His concern: a definition that obscures human accountability "will carry that blind spot into the policies it informs."
• Drawing on research in Brazil and the US, Marcelle Chagas, founder of GERATE Lab, argues digital policies fail when they treat communities as end users and ignore the local institutions that already organize trust. "Without trust, participation becomes decoration." Communities must be honestly engaged in AI governance, she says.
• Relatedly, Joyce Hakmeh, an associate fellow at Chatham House and a senior associate at Oxford Information Labs, and Lea Kaspar, executive director of Global Partners Digital (GPD), warn that 50 of 92 organizations that applied to join the UN's new cybersecurity process were blocked this year—risking rules that lose touch with the expertise they depend on. With the body convening next week, a key question is whether this will become the norm.

AI agents

• “If recommendation algorithms gave online platforms the ability to control what people see online, agentic AI offers them an opportunity to control what people do.” Rutgers law professor and Institute for Information Policy & Law (RIIPL) co-director Ellen P. Goodmanconsiders Sen. Mark Warner's draft AI AGENT Act and details obstacles to its implementation.
• Madhulika Srikumar, AI Safety Program lead at the Partnership on AI, and Vinh Nguyen, senior fellow for AI at the Council on Foreign Relations, argue regulators and developers agree AI agents must be monitored in real time—but the infrastructure to do it "doesn't exist yet in any mature form for the volume and complexity at which agents are being deployed."

AI and labor

• South Korean unions are demanding "not a single robot" on factory floors without worker consent. Tech Policy Press fellow Lam Le reports on how Asia's labor movements are fighting—and losing—ground as AI reshapes work.
• When AI hiring tools go wrong, vendors say the employer decides. Employers say the system decided. Samonne Montgomery, who holds a PhD in Planning & Public Policy from Rutgers University, writes that these are not opposite arguments—they are the same institutional dodge. “Oversight is real only when someone can challenge the system, override it, or explain it. A human who can only approve is evidence of process, not accountability,” she writes.
• Dr. Velislava Hillman, founder and director of EDDS Institute, and Dr. Peter G. Kirchschlaeger, professor of ethics and director of the Institute of Social Ethics ISE at the University of Lucerne, argue that as AI erodes job markets, it exposes the reality that the crisis facing young people "was never primarily technological, but political." Time to rethink what education is for, they say.

AI chatbots and safety

• A ban on romantic chatbots omits the general purpose AIs that teens actually confide in. Andrew McStay, professor of technology and society at Bangor University, writes that defining ‘intimate functionality’ in relation to emulated empathy will help. “‘Emulated empathy’ is technology engineered to display the appearance of strong empathy while possessing only the weak kind, and that gap—a system built to look as though it cares, which it does not—is what an intimate functionality is,” he writes.
• Voters are already asking AI chatbots how to vote. Marie Erikson, an independent writer interested in responsible tech and governance, says "Some chatbot regulation will support democracy. Too much will stifle it." Where's the line?
• Taposh Dutta Roy, director of innovation and AI at Kaiser Permanente and a bioethics affiliate at Harvard Medical School, argues the central ethical problem for mental health chatbots isn't alignment with human values—it's whether the institutions governing them are legitimate enough to decide whose risks matter.

EU digital regulation

• An EU expert panel has handed Brussels its most ambitious child safety blueprint yet: make platforms prove they are safe before children can reach them. Ramsha Jahangir, deputy editor at Tech Policy Press, reports on what it means and what comes next.
• Konrad Kollnig, an assistant professor at Maastricht University’s Law & Tech Lab, explores why EU member states are blocking the Digital Omnibus proposal to replace cookie banners with browser-level consent signals—and why the feared advertising revenue collapse likely won't materialize.
• The cost of building Europe's own frontier AI capacity may be very large. Igor Sevenard, Ph.D. Candidate at the University of Duisburg-Essen, and Richard J. Cook, associate professor of international relations at Nankai University, argue the cost of not building it may be even larger, and grow every year as the capability gap widens.

Censorship and surveillance in India

• Indian legal doctrine has distinguished "law and order" from "public order" for decades. Sarthak Gupta, New Delhi-based advocate and Researcher at Columbia University, argues the government’s June Telegram ban reclassified exam fraud as the latter for the first time—with substantial implications for the future.
• Facial recognition technology supplied by a Spanish firm and largely banned in the EU is running on around 4,000 cameras across India, report Nico Schmidt, Mayank Aggarwal, and Ignacio Carrascón for an investigative piece published in collaboration with Investigate Europe. “According to four leading legal scholars who specialize in EU artificial intelligence and biometrics law, two of these deployments would be deemed unlawful if they operated inside the European Union: the system on Indian Railways' Eastern Region and a city-wide surveillance program in Ahmedabad.

There are multiple pieces on other distinct topics that I recommend:

• Data sovereignty: Data sovereignty is not synonymous with isolationism, writes Luca Belli, professor of Digital Governance and Regulation at Fundação Getulio Vargas (FGV) Law School. He frames it as the capacity to know how and why data is processed, to extract value from it fairly, and to regulate those processes effectively.
• US science and technology policy: Using an open-source pipeline to read every comment on OMB's proposed grants rule that could have significant impact on US science and technology research funding, Christopher Steven Marcum, senior advisor on federal data policy at the Federation of American Scientists and a senior fellow at the Data Foundation, and Abigail Haddad, an independent data engineer, find the verdict "clear and lopsided": ~94% oppose it.
• Review bombing: Nearly all platforms that host consumer reviews have instituted new policies to govern reviews in response to the growth of ‘review bombing’ in the last decade. Andrew Wirzburger, who works at the Inter-university Consortium for Political and Social Research, part of the University of Michigan, argues that by classifying review bombing as "inauthentic" or "manipulative" and threatening to ban users for engaging in it, platforms have broken up "digital channels of dissent to ensure smoother consumer experiences."
• AI and copyright: When AI can instantly mimic creative work, "it's hard to imagine many current markets for creative human expression surviving," write Thomas W. Malone, Patrick J. McGovern (1959) Professor of Management at the MIT Sloan School of Management and the founding director of the MIT Center for Collective Intelligence, and Frank Pasquale, professor of law at Cornell Tech and Cornell Law School. Their answer: give creators a "learnright" and make AI companies pay to train.
• China vs US AI policy: Both the US and China are moving toward banning each other's most capable AI models. Georgetown University adjunct professor Mark MacCarthy argues these bans would be costly—and that a cooperative safety regime offers a better path.

What we’re watching

Despite Elon Musk’s vociferous criticism of the findings against X under the Digital Services Act, last week the European Commission accepted X’s plan to address the deficiencies that led to a €120 million fine.

And late Friday, Bloomberg reported that “The Trump administration is considering plans for an independent regulator to vet the safety of artificial intelligence models with industry input, after Silicon Valley leaders complained about the ad-hoc nature of recent US moves to slow the release of cutting-edge AI systems.”

We hope to receive perspectives on these developments and more in the week ahead. Learn more about our contributor’s guidelines here.

Until next week, enjoy the beautiful game!

-Justin