Good Afternoon:
Wednesday on #DogShirtTV, the estimable Holly Berkley Fletcher and I discussed the absolute mess of a failed execution in Tennessee, and the estimable Julia Curlee jumped in to talk about a major counterespionage failure in Germany:
Thursday on #DogShirtTV, the viola. Also, political prophylaxis and people who tour hell:
Friday on #DogShirtTV, the extremely estimable Ada Palmer came on to tell us about a new Hittite archive and the collapse of medieval Italian republics:
Monday on #DogShirtTV, the estimable Holly Berkley Fletcher visits with Boomers and argues with communists:
Yesterday on #DogShirtTV, the estimable Scott Shapiro came on to tell us about computational jurisprudence:
And today on #DogShirtTV, I was alone and bereft of cohosts, but the Greek Chorus rescued me:
Recently On Lawfare
Compiled by the estimable Athena Smith
The Law Can’t Keep Up With Flock’s AI Surveillance
Matthew Cavedon and Laura Bondank-Harmon describe the apparatus of Flock’s artificial intelligence (AI)-enabled surveillance tools and other technologies, such as license readers, drones, and smart cameras, and note that the Supreme Court has yet to determine whether many of these practices qualify as searches or seizures under the Fourth Amendment. The authors recommend policy changes—such as search warrants and ending qualified immunity—in combination with enforcement by the courts to protect civil liberties.
Without civil liberty protections, these methods may even be used against critics of AI surveillance. Flock’s chief executive Garrett Langley recently called the company’s opponents “terroristic.” Joint intelligence centers run by federal, state, and local law enforcement are monitoring social media accounts critical of AI-enhanced surveillance. After Rhode Island police officer Noel Pichardo criticized his department’s use of Flock cameras, “he was subjected to five internal affairs investigations in less than two years.” Policymakers should address the risks faced by citizens critical of AI surveillance.
How to Make an AI Deal With China: Trade Throttling for Pacing
Simon Goldstein and Peter N. Salib argue that, because concerns about Chinese competition are common objections to pausing AI research, any plan to pace AI development must include a deal China may accept. Goldstein and Salib explain that, to pace frontier AI development, the U.S. must trade its policy of limiting China through chip restrictions for a mutual agreement between the two countries to slow AI development.
We recognize that this suggestion will be met with significant resistance. Absent throttling, Chinese AI systems could quickly catch up to American capabilities, and could even take the lead. However, we think that accepting U.S.-China parity is likely the price of doing an AI safety deal. Neither country seems like it would be willing to accept the prospect of the other running away with the AI race.
Citizens’ Assemblies in AI Governance: An Interview With Audrey Tang
Kevin Frazier interviews Audrey Tang about the use of citizens’ assemblies in artificial intelligence (AI) governance. Tang argues that assemblies should name up front who must act on their recommendations, and that frontier AI labs should fund community evaluations and publish how they handle public input. Tang also explains that, although AI can automate clerical work, substantive work—the recruitment of participants, decision-making, and the response to an assembly—should be handled by humans.
Lawmakers should regulate outcomes rather than the specifics of today’s models, and that is precisely the layer where an assembly can make a difference. Which outcomes a community wants, where a data centre may be sited, who must answer when a system fails: these questions do not change every week. A model version changes in weeks. The values and the tests a community sets for it should hold for years and be revised on purpose. Deliberation is slow in the way a constitution is slow.
A Case Study in Pivoting: Investigating Hacked and Leaked Data on a Budget
Ceren Fitoz describes how her team at the University of California, Berkeley’s Human Rights Center analyzed six terabytes of leaked internal Mexican military emails for an investigation into the targeting of journalists and human rights defenders. Fitoz illustrates how, despite limited resources and technical setbacks, the group examined this data using network analysis tools; patterns in the military’s emails, signatures, and file paths; and an AI-designed search tool.
The takeaway here for any human rights research group or investigative team is to take the theory and put it into practice as much as possible; to try to handle complicated data problems, even when the team is learning as they go, as we did.
Our work is driven to expose the use of digital technologies in grave human rights violations. As bad actors harness sophisticated tools such as zero-click spyware or mass surveillance, we need human rights research to match the pace and creativity of these technology-enabled crimes. This investigation shows what we accomplished with little infrastructure; imagine what is possible with the right investment.
The AI Model Distillation Paradox
Lisa Klaassen argues that the U.S. government defends American artificial intelligence labs for training models on copyrighted work while branding Chinese labs a national security threat for distilling American models. Klaassen explains that these two positions cannot both be true, and Washington has yet to settle whether model outputs are legally protected.
Both propositions cannot hold simultaneously. Either the outputs of these systems are protectable under the law, in which case the writers, musicians, artists, and programmers whose work American models were built on can hold a claim that the Justice Department has denied. Or they are not, in which case the six Chinese laboratories accused of “malicious and targeted” distillation attacks have not done anything that Elon Musk has not freely admitted to under oath, without consequence.
The Cyber Corps: Ukraine’s Least Understood Cyber Actor
Stefan Soesanto describes connections between Ukraine’s military intelligence service (GUR) and a hacktivist group calling itself the Cyber Corps. Soesanto outlines a pattern of circumstantial evidence—linking the GUR to the Cyber Corps—including the Cyber Corps’ selective claiming of GUR-linked operations and its absorption of members of the volunteer IT Army of Ukraine, that collectively suggests a relationship between the two organizations.
Taken individually, each observation laid out in this article is open to plausible alternative explanations. Collectively, however, the evidence paints a different picture. The Cyber Corps consistently claimed a defined subset of GUR-linked cyber operations, published operational details that extended beyond official GUR information releases, repeatedly overlapped operationally with the GUR and the IT Army, was linked to GUR operations in multiple Ukrainian media reports citing intelligence sources, and ultimately absorbed experienced IT Army members.
Why the U.S. Military Has Never Threatened Civilian Governance
Loren Voss reviews Kori Schake’s book “The State and the Soldier,” in which Schake argued that the U.S. military has never posed a threat to civilian governance. Voss highlights the author’s historical account of military officers acting politically and the book’s conclusion that these examples are rare due to strong military professionalism. Voss argues that military professionalism alone cannot sustain healthy civil-military relations and notes how civilian leaders must also be constrained in how they exert power over the military.
Schake argues that despite the founders’ fears about the risks associated with a standing army, the U.S. military has never posed a threat to civilian governance. Throughout the book, she uses historical examples to explain how the military professionalism that so worried the founding generation has instead become a bulwark of American democracy. Her history effectively demonstrates the importance of military professionalism but also raises the question: What obligations do civilian leaders have to ensure the military does not pose a threat to democracy?
Dispatch: In ‘Fake News’ Ban Hearing, Trump Is His Own DOJ’s Worst Enemy
Molly Roberts describes the events in the courtroom during Cable News Network, Inc. v. Trump, a hearing to determine whether the temporary restraining order U.S. District Judge Timothy Kelly granted CNN following the Trump administration’s revocation of their White House access should be converted to a preliminary injunction. The plaintiffs argued the ban was viewpoint discrimination, while the Justice Department explained the plaintiffs had failed to contest by their deadline and were on notice not to publish classified information.
Boutrous now returns to the falsity question, this time with real force. The Supreme Court has never said that the First Amendment’s speech protections evaporate because of fraud, he says. Quite the contrary: Justice Anthony Kennedy, in his 2012 plurality opinion in United States v. Alvarez striking down the Stolen Valor Act of 2005 that criminalized falsely claiming to have received military honors, referenced “1984”: “Our constitutional tradition stands against the idea that we need Oceania’s Ministry of Truth.”
What Washington Got Wrong About the Houthis—and What It Must Do Now
Nadwa Al-Dawsari argues that Washington and its regional allies should not see the Houthis as simply an Iranian proxy but instead address the regenerative roots of the Houthi resistance. Due to the Houthis’ recent seizure of Yemen’s Red Sea coast—and the internationally recognized government’s Operation Dawn of Yemen to retake Houthi-held territory—Al-Dawsari notes that diplomacy and military action should support a broader goal of reducing the Houthis’ territorial, economic, and military base.
The experience of the past decade points to a basic conclusion: The United States cannot bomb its way out of the Houthi problem, but diplomacy without leverage will not solve it either. The missing element has been a strategy aimed at the source of Houthi resilience—the territorial base that allows them to recruit fighters, generate revenue, control ports and smuggling routes, manufacture and store weapons, and rebuild capabilities degraded by air strikes and sanctions.
The Many Open Questions About the Trump Administration’s New De Facto AI Model Licensing System
Adam Conner describes how President Trump’s June 2026 Executive Order 14409 on Promoting Advanced Artificial Intelligence (AI) Innovation and Security raises many questions about the text of its AI framework, which designates a threshold for covered frontier models but has been kept secret from the public. Conner explains that Trump’s sole discretion over the release of frontier U.S. AI models—without authorization from Congress—reduces transparency and safety around AI.
But it is worth stepping back and asking, What is the intent of the administration’s actions? The June AI EO was entitled “Promoting Advanced Artificial Intelligence Innovation and Security.” A secret de facto licensing regime is possibly the worst way to promote AI innovation and security. The secrecy is counterproductive because it undermines any ability to assure the public of the safety or security of new AI models at a time of growing concern, particularly as the government appears to have the opportunity to misuse unreleased models for its own purpose. Keeping this framework secret does nothing to reassure the public, other AI developers, Congress, U.S. allies, or even the nation’s enemies.
The Mercenary Exception That Never Existed
Lucas Gualtieri examines the Brazil’s Superior Court of Justice’s (STJ) ruling in the case of Lucas Passos Lima, a Brazilian national who surveyed synagogues for Hezbollah. The STJ reverses an earlier judgement by the Regional Federal Court of the Sixth Circuit, holding that one cannot escape liability for terrorism just because there were monetary incentives behind an action. The reversal signals an opportunity for greater international cooperation to prosecute the transnational networks used by Iran and Hezbollah.
Legal systems that focus only on the inner beliefs of the paid operative may miss the organizational relationship that gives their conduct meaning. The relevant distinction is not between ideologues and mercenaries. It is between a person who unwittingly provides an ordinary service and one who knowingly and voluntarily contributes to a terrorist plan. Payment may be evidence of recruitment, task allocation, and control. It should not become an immunity mechanism.
Podcasts
On Tuesday’s Lawfare Daily, Daniel Byman sits down with Peter Bergen to discuss counterterrorism in the 25 years after 9/11 and the many mistakes the United States and al-Qaeda made, as well as some of their surprising successes. Bergen also draws upon past lessons to understand the war in Iran and how the U.S. could better respond to a future terrorist attack.
On Scaling Laws, Kevin Frazier sits down with Hélène Landemore and examines who should decide the uses for AI. The pair discuss why decisions about AI’s purpose, limits, and governance should not be left solely to elected officials, technical experts, and the companies building the systems.
On Wednesday’s Lawfare Daily, Tyler McBrien sits down with Akbar Shahid Ahmed to discuss Ahmed’s new book “Crossing the Red Line: Biden, His Advisors, and Israel’s War in Gaza.” The two discuss the top officials in the administration who drove President Biden’s Gaza policy, how the war quickly became a moral and strategic disaster, and what the heavy civilian death toll might augur for the future.
On Thursday’s Lawfare Daily, Julia Curlee sits down with Shawn Howard and Ariane Tabatabai to consider how the Gulf states are coping with a war they didn’t start: the Iran War. They discussed why U.S. protection has made the Gulf states targets, how Tehran sees them, Iran’s strikes on the mediators, the possibility of a United Arab Emirates side deal, and what a post-midterm return to U.S. bombing would mean for the region.
On Rational Security, Scott Anderson sits down with me, Molly Roberts, and Molly Reynolds to talk through the week’s big national security news, including the White House’s pocket rescission of federal funding appropriated by Congress, the Trump administration’s efforts to reshape U.S. elections, and an incident of plague at a biological research institute in Siberia.
On Friday’s Lawfare Daily, Scott Anderson sits down with Omar Mohammed, Doug Silliman, and Victoria Taylor to discuss the recent withdrawal of U.S. military forces from Iraq 23 years after the U.S.-led invasion and occupation. The group considers whether the current Iraqi government is up to preventing a resurgence of ISIS and what lessons can be drawn from this experience and prior withdrawal in 2011.
Videos
On Friday, Oct. 9, at 4 p.m. ET, I sat down with Molly Roberts, Anna Bower, Roger Parloff, and Eric Columbus to discuss the preliminary injunction hearing in CNN’s lawsuit against the Trump administration; oral arguments in the Harvard University lawsuits over grants being cut by the administration and new visa restrictions; court-appointed U.S. attorney Roger Rogoff agreeing to stay out of office until the appeal is decided; and more.
Today’s #BeastOfTheDay is the crane, seen here having unreasonably long legs:
In honor of today’s Beast, stretch.
Tell Me Something Interesting
Last week on #DogShirtTV, there was some discussion about the uses and abuses of the phrases “yeah no” and “no yeah.” I—EJ Wittes—am a frequent user of “yeah no,” but I hadn’t given it much thought. Fortunately for me, other people absolutely have.