The Washington Roundtable Live: Can D.C. Get a Grip on A.I.?

Summary of The Washington Roundtable Live: Can D.C. Get a Grip on A.I.?

by The New Yorker

56m•September 25, 2026
Listen to Episode

Overview of The Washington Roundtable Live: Can D.C. Get a Grip on A.I.?

This live episode of The Political Scene from The New Yorker was recorded in Madison, Wisconsin at the Cap Times Idea Fest and centered on two major themes: the Trump White House’s escalating fight with the press and the growing political urgency around artificial intelligence. Jane Mayer, Susan Glasser, and Evan Osnos argued that AI is no longer a niche tech issue—it is becoming a national security, economic, and democratic governance crisis, with implications for U.S.-China relations, data centers, state politics, and the future of regulation.

Main Topics Discussed

Press freedom and Trump’s confrontation with the media

  • The conversation opened with the White House banning major news organizations and the broader attack on press access.
  • The hosts praised newsroom solidarity, especially Condé Nast’s amicus brief supporting affected outlets.
  • They argued that Trump has repeatedly tried to turn journalists into the story, while the press has generally tried to avoid taking the bait.
  • The panel noted that Trump’s media strategy depends on controlling the camera, the message, and the channel of distribution.

AI as a political and national security issue

  • The central topic was whether governments can respond meaningfully to AI before it becomes uncontrollable.
  • The hosts framed AI as more than an economic innovation:
    • a geopolitical competition between the U.S. and China,
    • a tool for authoritarian surveillance,
    • a potential source of bioweapons, drone attacks, and synthetic-virus threats,
    • and a disruption to labor, infrastructure, and public trust.

China, Trump, and the limits of diplomacy

  • Evan Osnos described China’s deep investment in AI as a control system and governance tool.
  • The hosts were skeptical that the Trump-Xi meeting would produce a real agreement, calling Trump’s approach too personal, transactional, and unserious for the stakes involved.
  • They noted that meaningful international AI rules would require both U.S. and Chinese participation, even if a full “arms control” style pact is unlikely soon.

Data centers as a new grassroots political flashpoint

  • The conversation emphasized how opposition to data centers is becoming a surprising local issue, especially in swing states like Wisconsin.
  • The hosts noted rising resistance across the country, including in states where people feel large tech projects are being imposed without consent.
  • They connected this backlash to broader frustration with inequality, infrastructure, and the sense that governments can mobilize instantly for corporate priorities but not for public needs.

Key Takeaways

  • AI is becoming a mainstream political issue. The hosts argued the “honeymoon” phase is over; public concern is shifting from optimism to alarm.
  • Congress is unlikely to lead without a crisis. They repeatedly suggested Washington is too dysfunctional and too influenced by money to regulate effectively on its own.
  • The political spectrum is oddly converging. Figures as different as Bernie Sanders and Steve Bannon are reacting to AI and data centers through a shared distrust of concentrated power.
  • The real threat may be domestic as much as foreign. Several comments stressed that the biggest obstacle may be the U.S. political system itself, not just China.
  • Regulation is being reframed as necessary, not anti-business. Susan Glasser and others argued that regulation can protect innovation and credibility, rather than stifle it.

Audience Q&A Highlights

Can AI regulation become an international project?

  • A question from the audience asked whether other countries would be equally concerned about AI-driven catastrophe.
  • The panel agreed that this is one area where shared vulnerability could create room for international negotiation.
  • Osnos noted that Chinese officials are increasingly treating AI as a national security issue, not just an economic opportunity.

Why don’t reporters push back harder on Trump?

  • Another audience member criticized journalists for not confronting Trump more directly.
  • The panel responded that press strategy has centered on solidarity and not becoming the story themselves.
  • They emphasized that Trump often uses outrage as a diversion from other problems.

Who should people look to for leadership?

  • The hosts pointed to a few lawmakers, especially Mark Warner and bipartisan efforts from Richard Blumenthal and Josh Hawley, but said Congress is still largely reactive.
  • They also stressed that one in four lobbyists in Washington is now working on AI, showing the scale of industry pressure.

Notable Insights

  • “The greatest geopolitical threat in the world today is the United States.” — a blunt takeaway from the discussion of instability and weak governance.
  • AI was repeatedly compared to nuclear weapons, but with one major difference: it is being developed largely by private companies.
  • The hosts warned that the next major AI shock could be a “holy heck moment”—a disaster big enough to force action after the fact.
  • They also argued that by 2028, AI will likely be a major presidential campaign issue, especially as public fear overtakes early optimism.

Bottom Line

The episode presented AI as a fast-approaching political reckoning rather than a distant tech debate. The hosts argued that Washington is behind, Congress is weak, the lobbyists are flooding the system, and the public is only beginning to grasp the scale of the danger. Their core message: AI regulation, press freedom, and democratic accountability are all colliding at once—and the U.S. is not yet equipped to handle any of them well.