If AI Breaks the Internet, Should We Abandon It?

Good morning!

I saw so many Tech Policy Press readers and contributors at the Mozilla Festival in Barcelona last week, and yet I also regret missing so many others. There were simply too many people and ideas and too few hours.

The theme of the festival was “wilding” — no, not that “wilding” — and “what it means to let our digital worlds grow wilder, more diverse, and more alive.” There were discussions on a range of issues that connected to the theme. One of the most interesting tensions in the program was over the opportunity around AI, and the threat it poses.

We have two special pieces on Tech Policy Press this week that connect to the theme.

The first is a podcast that asks, should we abandon the internet? The question is intended as a provocation. After the Privacy Center at Georgetown Law’s Life After Data conference, I asked newsletter readers for two-minute voice memos, and responses ranged from logging off for good to rebuilding the internet on new protocols. Hear a selection of the responses from Douglas Rushkoff, Anika Collier Navaroli, Bridget Todd, Konstantinos Komaitis, Caddie Alford and her VCU students, Janet Vertesi, Alice Marwick, Liz Orembo, Robin Berjon and B Cavello in this week’s episode, produced by Rebecca Rand.

The second is a thoughtful consideration of what to do about the threat that AI poses to the commons. Public Knowledge’s Lisa Macpherson and Creative Commons’ Anna Tumadóttir argue AI has broken the open web’s old bargain and propose principles for “a new internet bargain that works for everyone” while preserving fair use and protecting the knowledge commons. Whether you agree with their conclusions or not, this careful argument is worth your time.

Read on for more from our contributors this week!

Special investigation into data centers in Europe

• A five-month investigation by Investigate Europe in partnership with Tech Policy Press and media partners in nine countries reveals the harsh realities facing some working on the continent's vast AI infrastructure developments. Thanks to Mei-Ling McNamara and Jan Daalder for their dogged reporting on this investigation, which has already produced results. Just days before the investigation was published, a landmark collective agreement for better pay and benefits was brokered between migrant workers and a large Irish construction firm building a Microsoft data center in Sweden.

Platforms and democracy

• Brazil re-elected Lula in a historic comeback. He’s already signaling that social media regulation will be a top priority, writes Tech Policy Press fellow Tatiana Dias. “In this election, Big Tech became a symbol of US imperialism over Brazil, and became a mainstream concern,” she writes.
• Whether Meta favored Flávio Bolsonaro in Brazil’s election will “need to wait for new revelations, research, or, ideally, a full investigation,” write Clara Iglesias Keller of the Weizenbaum Institute, Robert Gorwa of the Berlin Social Science Center, and João C. Magalhães of the University of Manchester. But they argue the episode offers lessons for how platform governance is changing.
• Political influencers operate within a “complex, opaque multimillion industry,” write Isabel Linzer of the Center for Democracy & Technology and Zoe Tzanis and Samuel Woolley at the Communication Technology Research Lab (CTRL) at the University of Pittsburgh. They lay out how money, platforms and regulations (or lack thereof) shape it.
• Priscilla Yoon, graduate student at American University, argues prediction markets like Kalshi and Polymarket are expanding and restricting democratic participation at the same time by turning politics into a commodity, with risks that range from foreign interference to eroded trust in elections.

AI governance and AI politics

• The Pentagon’s AI dependencies stem from “a decades-long shift toward privatization of the US military’s weaponry and technology needs,” write Amos Toh of the Brennan Center for Justice and Margaret Mullins of the Vanderbilt Policy Accelerator.
• Mark MacCarthy of Georgetown University warns “the overexpansion of AI data centers is so large that the contraction is likely to have substantial macroeconomic consequences.” He says Congress should still act on ratepayers, safety, liability and AI misinformation.
• Vasant Dhar and João Sedoc of NYU Stern School of Business and the Center for Data Science at NYU argue that the answer to AI’s alignment problem lies not in hard guarantees from the technology, but in clear expectations and guardrails around its “use and consequences.”
• Imane Bello, international AI governance lead at the Future of Life Institute, argues the US has quietly redefined “superintelligence” to mean just “AI,” erasing any measurable threshold. Her case: Canada and the EU should set their own capability thresholds — and demand to see the frontier models they can’t access.
• Aggressive AI scraping “is a real problem that deserves a real solution,” write Lila Bailey of the Internet Archive and Tess Vartanian of the Center for Democracy & Technology. But they argue “stealth crawler” bills target anonymity, not the harms.
• California and the EU now require AI developers to summarize their training data. But “uncovering the contents of training datasets alone is insufficient to enable meaningful accountability,” writes Vanderbilt’s Will Orr, who outlines four ways to make dataset transparency meaningful.
• Rachel A. George, a Stanford University lecturer on international relations, argues AI partnerships between wealthy donors and poorer countries risk becoming a new form of tied aid that locks recipients into donors’ priorities, standards, and suppliers, and says disclosure and credible exit paths can prevent that.
• The UK-US Tech Prosperity Deal has returned to the negotiating table, but its revival remains partial. Researcher Megan Kirkwood writes that renewed talks raise questions about whether the UK will trade away digital regulatory powers in pursuit of AI and closer ties with Washington.
• New DDIA polling of 1,500 Latino registered voters finds 78% see “a pressing need for stricter rules and oversight of AI,” including 72% of Republicans. Roberta Braga of the Digital Democracy Institute of the Americas lays out what policymakers can take from it.

A review of Alex Gibney’s ‘Musk’

• Elon Musk was introduced at a MAGA rally as “the greatest capitalist in history.” Reviewing Alex Gibney’s “Musk,” Paul M. Barrett, former deputy director of the Center for Business and Human Rights at NYU’s Stern School of Business, writes that Musk instead combined “an uncanny ability to hoover up government subsidies and contracts with a P.T. Barnum-level talent for hucksterism.”

What we’re watching

• This week, voters in the US go to the polls to cast their ballots in the midterm elections. The outcomes will shape the direction of tech policy, of investigations in Congress, of policy priorities in state houses, and of who gets heard in the legislative and political battles in the weeks and months ahead. We invite your perspectives on these matters, and more.

One final note: our survey is still open — we’re looking for feedback! It’s just a few short, open-ended questions. Please reply!

I wish you the best for the week ahead.

Justin