The Lawfare Podcast · The Lawfare Institute

Rational Security: The “Authentic Flavors, Real Fruit” Edition

March 26, 2026·1 hr 7 min·5 clips

Meta's internal whistleblowers testified about over half a million inappropriate actions with minors happening every single day.

This episode of Rational Security, hosted by Scott Anderson with Renee DiResta, Tyler McBrien, and Molly Roberts, covers three national security developments in depth. The first and longest segment examines two recent jury verdicts against Meta in New Mexico and Los Angeles. The New Mexico case centered on Operation Metafile, where undercover agents posing as 12- and 13-year-old girls found that Meta's platform steered adult followers toward the fake child accounts and then offered the accounts tips on monetizing their following rather than flagging suspicious activity. Internal whistleblowers testified that over half a million inappropriate interactions with minors occurred daily. The jury awarded $375 million in damages. The Los Angeles case focused on whether Facebook and YouTube's design features, particularly endless scroll, created addictive experiences harmful to an individual plaintiff. DiResta frames both cases as a shift from adjudicating content disputes to treating platform conduct and product design as the source of harm, drawing comparisons to tobacco industry litigation. The panel debates where Section 230 immunity ends when algorithms actively steer users rather than passively hosting content, noting that 41 percent of Meta's widely viewed content now comes from accounts users do not follow, up from 26 percent two years prior. They discuss implications for AI liability, noting that AI chatbots directing users toward self-harm present even clearer attribution of harm than algorithmic feeds. The second segment covers the legislative battle over the SAVE America Act, which contains election security provisions including proof-of-citizenship voting requirements that Trump views as vindication of his 2020 election claims. Senate Republicans resisted blowing up the filibuster to pass it, instead reaching a compromise to fund non-controversial DHS operations through regular order while routing ICE funding and select SAVE Act provisions through reconciliation on a party-line vote. The panel notes that many of these voter-restriction proposals predate Trump and are rooted in longstanding myths about voter fraud, and that 2024 election results showed Trump gained among demographics most likely to be affected by stricter ID requirements, complicating the partisan calculus. The third segment addresses the explosion of open-source intelligence accounts on social media platforms like X, which have gained millions of followers by monitoring events from the Russia-Ukraine war to natural disasters. The panel discusses how OSINT intersects with prediction markets like Polymarket and Kalshi, where participants use open-source data for advantage and sometimes manipulation.

As heard by us

A careful look at where chatbot risk starts to look like platform liability.

The Lawfare crew uses this Rational Security installment to wrestle with a messy legal question: when a chatbot starts drifting toward self-harm, where does responsibility actually land?

Read the full review in PlayNext →

Why you'd press play

The legal line between chatbot coaching and liability gets real here.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from The Lawfare Podcast