Rational Security – “Sweet Dreams Are Made of Cheese” Edition

Podcast by The Lawfare Institute | Episode Date: October 1, 2025
Panel: Scott R. Anderson (Host), Anna Bauer, Alan Rozenshtein

Episode Overview

This week, the Rational Security team returns to debate and dissect a jam-packed national security news cycle, focusing on three significant topics:

1. The controversial DOJ indictment of former FBI Director James Comey,
2. California’s new law restricting law enforcement masking (especially targeting ICE agents),
3. California's new AI safety and transparency law.

Interspersed with the signature wit and self-deprecating humor of the panel, the discussion pushes deep into the legal, constitutional, and political implications of these stories.

Key Discussion Points & Insights

1. The Indictment of James Comey: DOJ Under Pressure

[04:45–40:28]

Background

• Trump administration Justice Department has indicted former FBI Director James Comey on charges of obstructing a congressional investigation and making false statements to Congress ([06:57]).
• The case follows years of Trump’s public attacks on Comey, ongoing since the Russia investigation. Recently, a top DOJ prosecutor in the Eastern District of Virginia, Eric Siebert, resigned reportedly after refusing to pursue charges against Comey and other perceived Trump enemies ([06:57–09:23]).
• Siebert was replaced by Lindsey Halligan, a former Trump defense attorney with no prosecutorial experience, who swiftly secured Comey’s indictment ([09:24–13:55]).

Details of the Indictment

• The indictment is sparse—just two short passages cover the alleged false statements by Comey, particularly referencing 2020 congressional testimony about authorizing anonymous sources ([09:40–13:55]).
• Much confusion remains over the specifics, including which statements and authorizations are at issue.

The Panel’s Take

• Procedural Anomalies & Legal Ethics:

  • The DOJ is acting “shockingly outside the norm,” pursuing the case despite career prosecutors and even AG Pam Bondi reportedly advising against it ([18:09–22:24]).
  • The indictment is unusually bare-bones for such a high-profile target; typically, a “speaking indictment” would be detailed to support credibility ([18:09–22:24]).
  • Alan blasts the prosecution as “the wholesale collapse of the Justice Department as a rule of law institution” ([18:17]), and Anna underscores that grand jury presentations are usually so one-sided that it is startling even two (possibly more) jurors opposed the indictment ([22:58–26:17]).
  • “I mean, I’ve managed to get an indictment before a grand jury. … It’s not that hard. So the failed, the no true bill is a big deal.” – Alan ([26:17])

• Comparisons & Whataboutism:

  • Alan criticizes both-sides rhetoric, noting the immense procedural and institutional differences between Trump indictments and Comey’s: “It cannot be emphasized enough how different the processes were…” ([27:45]).

• Vindictive Prosecution as a Defense:

  • Anna details that motions to dismiss for vindictive prosecution are usually an uphill battle, but this case—given Trump’s open calls for prosecution—may present the “strongest … in American history” ([29:56–35:25]).
  • Discovery over behind-the-scenes DOJ pressure could extend litigation for years ([35:25]).

• Potential Consequences

  • The panel expects Comey’s defense to file for vindictive prosecution, which, if successful, could have the case dismissed for procedural abuse rather than factual innocence ([35:25–38:07]).
  • “If on the one side you’ve got James Comey and his very experienced criminal defense and former federal prosecutors who are defending him, on the other side you’ve got Lindsey Halligan. And maybe nobody.” – Anna ([38:07])

Notable Quotes

• “The polite way of saying this is that this is very unusual. The more accurate way is that this is the wholesale collapse of the Justice Department as a rule of law institution.” – Alan ([18:17])
• "Trump got to express his will. Trump’s will triumphed today, as it were. … Halligan may not care because, you know, sure, she might get in bar discipline, but you don’t need to be a member of the bar to be a fourth Circuit judge." – Alan ([22:46])
• “Considering how one-sided this … grand jury presentation is, it is remarkable that at least two and potentially nine grand jurors did not find probable cause to indict Comey.” – Anna ([25:30])

2. California’s Law Limiting Police Masking—A Federalism Showdown

[41:41–63:14]

The Law’s Provisions

• New California law requires law enforcement agencies—including federal entities operating in the state—to publish policies limiting use of face masks ([41:53–43:54]).
• Target: ICE and federal agencies using masks during routine law enforcement, especially at protests.
• Civil and criminal penalties for violations, and agents exceeding their policy’s prescription risk loss of certain immunities ([41:53–43:54]).

Constitutional Debate: Supremacy Clause & Nullification

• Alan argues the law is “pretty obviously unconstitutional”—a violation of the Supremacy Clause: “This is just state nullification.” ([47:29])
• Points to a long history of federal authority trumping state attempts to regulate federal agents’ conduct ([45:21–47:45]).
• Anna and Scott push back: the law’s main effect is disclosure—not a direct bar to federal duty. It challenges the “necessity” of masking for federal duties, and masks are not the same as, for example, a requirement to have a state driver’s license for a postal worker ([48:00–52:22]).
• Scott sees it as a clever effort to force court clarity on the line between federal and state authority: “As a like, effective piece of legislation intended to force clarity on the legal limits of this and put pressure on the agency, it actually strikes me as like really cleverly designed.” ([55:57]).

Tensions & Tone

• Alan warns: “Once you go down this path, it gets ugly very quickly.” ([47:44])
• Scott: “I don’t disagree, but … there’s a lot of merit in … testing the limits of the federal authority. This is a good moment to do that.” ([62:49])

3. California’s New AI Safety Law: Transparency Over Prohibition

[63:20–75:23]

SB53 – Key Protections

• Sets up a public task force for “Cal Compute,” a state-run computing cluster—a “public option for computation” ([65:05–65:28]).
• Enhances whistleblower protections in major AI companies ([65:29–65:44]).
• Most notably: Major AI labs must publicly disclose their safety plans and risks. The requirement is not for having plans, but for publicizing the extent and content (even if they have none) ([66:07–67:51]).
• Only applies to the biggest companies ($500M+ annual revenue).

Industry Response

• Unlike previous, tougher proposals, this law was met with little or no resistance; firms like Anthropic openly endorsed it ([64:33–65:07]).

First Amendment Issues

• Alan voices concern about whether the compelled disclosures are “purely factual and uncontroversial,” the standard from Zauderer v. Office of Disciplinary Counsel ([67:51–70:07]).
• Uncertainty remains: Is forcing an AI company to estimate existential risk “purely factual”?

Broader Implications

• Should states like California regulate a fundamentally national/international issue like AI? The panel is mixed, but sees value in “starting the legislative process going” ([71:15]).
• Scott: “This is at least a real first step towards … consumer safety … so that consumer groups … can start saying … things to weigh in on these different tools” ([73:13]).

Memorable Quotes & Moments

• On Lawfare Team’s Attire:
  “It says, ‘Sweet dreams are made of cheese. Who am I to disabree?’ … You have to be over the age of 35 to like this shirt.” – Alan ([01:05])

• On Prosecution Practice:
  “Brought an indictment in a page? … Part of that you do it because it’s a show of strength. … I mean, we all had a good laugh as we were reading through … the indictment against Bob Menendez and the gold bars.” – Alan ([18:17])

• On Nullification Dangers:
  “This is just state nullification. … Once you go down this path, it gets ugly very quickly.” – Alan ([47:29])

• On AI Law’s Strategy:
  “It’s clever, right? I mean, it’s funny … forcing these entities … to be transparent about what they’re doing.” – Alan ([67:51])

• On AI’s Magic:
  “They’re kind of magic. Like I understand how they work, but I understand on a conceptual level … They’re black boxes.” – Scott ([72:11])

• Comedic Aside:
  “We actually replaced Scott with an AI voice model and video model months ago.” – Alan ([80:54])

Object Lessons

[75:37–87:25]

• Alan: Enthuses about “vibe coding” with the new Anthropic Claude 4.5 Sonnet model—a way AI is making coding fun, accessible, and more creative ([75:37–78:05]).
• Scott: Recommends Google’s NotebookLM for organizing, learning, and extracting insight from dense research—AI as a supercharged research assistant ([78:05–80:54]).
• Anna: Urges listeners to visit a local comedy club for laughter in dark times, and plugs Taylor Swift’s new album for a bit of millennial nostalgia ([81:17–86:10]).

Useful Timestamps

• Cheese T-shirt Banter: [00:33–04:15]
• Comey Indictment, Procedural & Legal Issues: [04:45–40:28]
  • Ethics, Grand Jury Math: [15:59–26:17]
  • Vindictive Prosecution: [29:56–38:07]
• California Masking Law, Federalism Debate: [41:41–63:14]
• California AI Safety Law: [63:20–75:23]
• Object Lessons (AI coding, research tools, comedy): [75:37–87:25]

Summary & Tone

This episode captures Rational Security’s signature: expert analysis with approachable humor and intellectual openness. The hosts maintain skepticism, a dash of outrage, and a willingness to challenge each other—mixing dad jokes, sports analogies, millennial references, and the occasional existential musings on AI. Policy buffs and general listeners alike will find this episode both entertaining and enlightening on some of the toughest legal questions confronting American democracy and technology in 2025.