Nine Things Platforms Could Do Now to Help Blunt Political Violence in the United States

Good morning!

As the United States approaches its 2026 midterm election cycle, the nation faces a volatile set of circumstances. Polarization and political violence have risen sharply: unprecedented levels of threats against public officials both on- and offline have coincided with a bout of assassination attempts and acts of targeted violence.

While perpetrators of violence are responsible for their own actions and factors well beyond social media contribute to the increasing polarization the US is experiencing, online platforms have enabled once fringe views to become more mainstream. They are tools for recruiting individuals; distributing messages and narratives intended to incite political violence; and organizing groups that may carry out violent acts. At the same time, online platforms' trust and safety policies and teams have been significantly degraded, as has the federal government's infrastructure for addressing foreign interference and election security.

In a new report published this week, Yaël Eisenstat, director of policy and impact at the Cybersafety Research Center and I argue the people who own and lead online platforms are responsible for the choices they make in designing and governing them—and those choices have real consequences. Platforms are unlikely to match the scale of their 2020 or 2024 preparations, but there are still steps they could take ahead of November. To that end, we offer a series of recommendations. The full set of nine recommendations—encompassing both immediate steps and broader design and product decisions required to reverse course toward a healthier and safer online environment—are spelled out in the report.

Yaël and I are grateful for the input and support of all the working group contributors to this report, including Jeff Allen, Integrity Institute; Luke Barnes, NYU Stern Center for Business & Human Rights; J.M. Berger, Center on Terrorism, Extremism, and Counterterrorism at the Middlebury Institute of International Studies; Dean Jackson, Tech Policy Press; Daniel Kreiss, Hussman School of Journalism and Media and the Center for Information, Technology, and Public Life at the University of North Carolina at Chapel Hill; Anika Collier Navaroli, Columbia University Graduate School of Journalism; Spencer Overton, Multiracial Democracy Project, George Washington University Law School; Katie A. Paul, Tech Transparency Project; Stephen Richer, former elected Maricopa County Recorder; and Dhanaraj Thakur, Multiracial Democracy Project, George Washington University Law School.

We had multiple pieces this week that connected to the themes of the report, including on extremism, the threat of state violence, and the problem of AI-generated falsehoods:

• Anti-tech extremism: Treating all opposition to data centers as a national security threat is the wrong move, says Jordyn Abrams, research fellow at the Program on Extremism at The George Washington University, who argues policymakers must address legitimate public grievances about AI—not just police them.
• Chilling of dissent: “Crackdowns on dissent aren’t a reflection of power—they’re a projection of power,” writes Jenna Ruddock, advocacy director at Free Press and Free Press Action. The Trump administration wants to deter any organizing efforts that effectively interrupt its agenda—using indictments to make a chilling example of a few—but using Signal, making zines and watching the watchers are all constitutionally protected activities.
• AI-generated text: The International Panel on the Information Environment’s (IPIE) Sebastián Valenzuela and Aliaksandr Herasimenka say lawmakers must recognize that the biggest generative-AI misinformation threat is not visual but textual. The science says giving people context before they see false content works best to address the problem.

A worrying precedent in India

• Can the state legally switch off an entire app under a law meant for blocking "specific information"? The Delhi High Court said yes. Amber Sinha, contributing editor at Tech Policy Press, unpacks why that reasoning should worry every platform in India. “The Telegram FZ order has effectively codified a dangerous precedent: in the digital age, the Indian state no longer needs to censor the message if it can simply turn off the medium,” he writes.
• India's Telegram block lasted six days; the power behind it doesn't expire, writes Tech Policy Press fellow Apar Gupta. The Delhi High Court let the government treat an entire platform — used by 150M Indians — as blockable "information."

Brussels watch: digital regulation in the EU

• The EU is debating a minimum age for social media — but MEPs want to know why existing DSA protections for minors aren't being enforced first, Ramsha Jahangir, deputy editor at Tech Policy Press, reports.
• The EU’s Code of Practice on Transparency of AI-Generated Content is now final — and Article 50 compliance kicks in Aug. 2. Natalia Garina, a legal researcher and consultant specializing in AI policy, digital rights, and freedom of expression, walks through what providers and deployers need to know and whether it will prove effective in practice.
• Google and Apple are framing the EU's DMA interoperability mandatesas privacy and security risks, writes George Colville of Open Markets Institute Europe. He argues these claims should be assessed by independent experts, not incumbents with conflicting interests. “Privacy and security are legitimate concerns, but tech giants appear to be taking advantage of this to defend their market power,” he writes.

A new prime minister in the UK (again)

• How will Andy Burnham approach tech policy if he becomes UK Prime Minister? Tech Policy Press fellow James Ball reads the clues: skepticism on surveillance, “not now” on digital ID, and likely continuity on Starmer’s under-16 social media ban.
• The ban proposed in the UK is a symbolic reset, writes Maeve Walsh of the Online Safety Act Network, but treating social media harms as a product safety failure—not just age-gating and hoping for the best—gets at the root, and it's already doable under the UK’s existing Online Safety Act.

All AI all the time

• Late Friday, the Commerce Department eased its block on Anthropic's Mythos, but UC Berkeley Risk and Security Lab's Andrew W. Reddieargues Secretary Howard Lutnick's letter leaves the hard question unanswered: in an age of cloud-hosted models, what even counts as an "export" when a capability crosses borders, but not the code?
• Carl Schonander, a global technology policy professional and former diplomat with the US State Department, and Mark MacCarthy, an adjunct professor at Georgetown University, argue Congress should pass AI laws affirming that companies are responsible for the products they produce, focused on catastrophic risk and consumer harm.
• The techlash needs a target close to home, say Arjun Hassard and Erin McElroy of the Anti-Eviction Lab, arguing tenant screening—with its algorithmic discrimination, junk fees, and lifelong digital stigma—is the bold, common-sense tech fight candidates should pick.
• AI wearables can capture your voice, image, and likeness without you ever knowing. Attorneys Jannet Riveros Gomez and Samantha Nicole Spector examine why the law hasn’t kept up.
• Max Belasco of UPTE UC Los Angeles explains how the University of California's IT workers added 2,100 members to union rolls—and won the right to bargain over how AI gets deployed. He explains how he and his colleagues knocked on hundreds of doors to build solidarity from the bottom up.

Special series: Agency, trust, and governance in the age of AI

This spring, the Berkman Klein Center for Internet & Society at Harvard University hosted an essay contest in collaboration with Tech Policy Press focused on the ways in which AI systems intersect with power and the public interest. Last week we published the remaining four pieces in the series, all of which connect to the themes above:

• Recent Harvard computer science graduate Hannah Kim reflects on attending college in the wake of ChatGPT, and what AI can't do for students: the work of becoming. “Just as students can still choose to wrestle with a problem before asking for the solution, they can choose to wrestle with a blank page before prompting a large language model primed with an ‘academic tone’ and ‘no em dashes’,” she writes.
• What happens to cities if AI hollows out the job market? Isabel N. Adler, an urban planning student at Harvard Graduate School of Design, issues a call to arms for urban planners: fight to build cities that value residents over capital accumulation for the wealthy few. Imagination and resistance are needed now, before the shock arrives, Adler argues.
• David Gantt, a J.D. candidate at Harvard Law School, supports Anthropic's red lines on surveillance and lethal autonomous weapons—but argues governing AI risks by contract produces "dangerously brittle" civil liberties protections that only public law can secure. “It is tempting, when facing an administration as reckless as the current one, to cheer any institution that stands in its way. But the right response to democratic dysfunction is not to cede governance to private corporations,” he writes.
• "A sequence of individually approved operations can produce one catastrophic outcome." Kida Chung-Ta Huang, a researcher at the Harvard AI and Robotics Lab, on why AI agent oversight fails—and what outcome-level governance would actually require.

Protecting free expression

• Treating surveillance as an isolated privacy problem is the wrong frame, says UN Special Rapporteur Gina Romero, who argues digital monitoring is dismantling the trust, funding, and institutional memory that civil society depends on to function.
• As governments rush to age-gate the internet, Electronic Frontier Foundation (EFF) senior counsel David Greene argued this week in Tech Policy Press that we're skipping a vital question: not "how do we age-gate while protecting privacy?" but "should we be doing this at all?"

Listen

We published two podcasts this weekend. Both discussions are concerned, in their own ways, with civil rights and equity issues:

• Digital equity: Access to affordable, reliable high-speed internet is a prerequisite for nearly every part of modern life, from finding work and finishing schoolwork to seeing a doctor or staying in touch with family. Yet millions of American households remain stranded on the wrong side of the digital divide. That's the starting point for "The Blueprint for Equitable Digital Participation," a report released in May from Public Knowledge, UnidosUS, and the National Digital Inclusion Alliance. Rather than beginning in Washington policy circles, this report centers the lived experiences of low- and moderate-income households to find out what's actually standing in their way and what should be done about it. I had the chance to dig into the findings with the report’s authors: Alisa Valentin, broadband policy director at Public Knowledge, and Claudia Ruiz, senior civil rights policy advisor at UnidosUS. Listen here.
• xAI in Memphis and Mississippi: Two years on, the story around xAI’s presence in Memphis continues to expand alongside the company’s growing footprint, with a second campus, Colossus II, across the state line in Southaven, Mississippi; a contested gray water recycling plant; an ever-rising count of gas turbines; multiple lawsuits; and communities in South Memphis and beyond still pressing for straight answers. One of the entities leading the way in Memphis is the Southern Environmental Law Center (SELC). The SELC has been a partner in litigation against xAI, gathering evidence on the company's gas turbines and pushing state regulators to hold the company accountable under the Clean Air Act. To learn more about SELC’s case and what it means for the broader fight for transparency and accountability amid the AI infrastructure boom, I spoke to Amanda Garcia, senior attorney and data center project leader in the Tennessee office of the Southern Environmental Law Center. Listen here.

What we’re watching

The debate over age restrictions continues to heat up around the world. We expect more developments on this and other child online safety matters in the days ahead, including related to the KIDS Act in the US, which could see a House vote this week. We invite analyses and perspectives on these issues.

I wish you the best for the week ahead!

-Justin