June 29th • 3h 26m
Shownotes
Every new episode of No Agenda is accompanied by a comprehensive list of shownotes curated by Adam while preparing for the show. Clips played by the hosts during the show can also be found here.
Big Tech AI and the Socials
Passkey additional info
Hello Adam,
While passwords are protected under
the Fifth Amendment since you can’t be forced to reveal what you know,
passkeys work differently. The user doesn’t actually know the passkey;
it’s stored and accessed with biometrics or a device PIN. Because of
that, I'm pretty sure that passkeys generally don’t get the same legal
protection against compelled use—similar to Face ID or fingerprints.
With more platforms pushing passkeys, it's an important distinction to keep in mind.
All the best,
Punched In The Podcast
Big Beautiful Bill AI restrictions and remedies
To address whether states can avoid the prohibition on making or enforcing laws or regulations restricting AI systems under **Section 43201(c)** of the “One Big Beautiful Bill Act” (H.R. 1), let’s analyze the provision and potential legal or practical avenues for states to retain regulatory authority. The prohibition, as outlined, imposes a 10-year moratorium on states enforcing laws or regulations that limit, restrict, or otherwise regulate artificial intelligence models, systems, or automated decision systems, with limited exceptions.
### Recap of Section 43201(c)
**Text:**
“Except as provided in paragraph (2), no State or political subdivision thereof may enforce, during the 10-year period beginning on the date of the enactment of this Act, any law or regulation limiting, restricting, or otherwise regulating artificial intelligence models, artificial intelligence systems, or automated decision systems.”
**Key Details:**
- **Exceptions in Paragraph (2):** States can enforce laws that “remove legal impediments” to AI deployment or align with federal requirements under the Act.
- **Criminal Penalty Exemption:** An amendment allows states to enforce laws carrying criminal penalties, exempting them from the moratorium.
- **Scope:** The moratorium applies universally to all states, regardless of whether they accept federal funding (e.g., for IT modernization).
- **Status:** As of June 28, 2025, H.R. 1 has passed the House (215-214 on May 22, 2025) and is under Senate consideration, potentially facing Byrd Rule challenges.
### Can States Avoid the Prohibition?
States have limited options to avoid or mitigate the prohibition if the bill becomes law, but several legal, practical, and strategic approaches could be considered. Below are the primary avenues:
#### 1. Leverage Existing Exceptions in the Bill
States can work within the bill’s exemptions to maintain some regulatory authority over AI systems:
- **Criminal Penalty Laws:**
- The House Rules Committee amendment exempts state laws with criminal penalties from the moratorium. States could frame AI-related regulations (e.g., against deepfakes, AI-driven fraud, or privacy violations) as criminal statutes to retain enforcement power.
- **Example:** A state could pass a law criminalizing the use of AI to create non-consensual deepfake content, which would be enforceable despite the moratorium.
- **Limitation:** This approach is narrow, as not all AI regulations (e.g., consumer protections, algorithmic transparency) can be effectively structured as criminal laws without stretching legal frameworks.
- **Paragraph (2) Exceptions:**
- States can enact laws that “remove legal impediments” to AI deployment or align with federal requirements. This could include regulations that facilitate AI use in ways consistent with federal goals (e.g., promoting cybersecurity in AI systems).
- **Example:** A state could require AI systems to meet federal cybersecurity standards, framing it as enabling rather than restricting AI deployment.
- **Limitation:** The scope of “removing legal impediments” is vague and may be subject to federal interpretation or legal challenges, limiting its practical utility.
#### 2. Legal Challenges to the Moratorium
States could challenge the moratorium’s constitutionality or legality, potentially delaying or overturning its application:
- **Federalism and Tenth Amendment Claims:**
- The Tenth Amendment reserves powers to states not delegated to the federal government. States could argue that the moratorium infringes on their sovereign authority to regulate technology within their borders, especially in areas like consumer protection, public safety, or privacy, which are traditionally state domains.
- **Precedent:** In cases like *Printz v. United States* (1997), the Supreme Court struck down federal laws that compelled state officials to enforce federal regulations, citing anti-commandeering principles. States could argue that the moratorium effectively commandeers their legislative authority by prohibiting state laws.
- **Challenge:** Congress has broad authority under the Commerce Clause to regulate interstate commerce, and AI’s national and global impact could justify federal preemption. Courts may uphold the moratorium if it’s deemed a valid exercise of federal power.
- **Preemption Doctrine:**
- States could argue that the moratorium’s preemption of state law is overly broad or vague, failing to provide clear federal standards for AI regulation. Courts sometimes strike down preemption clauses that lack specificity (e.g., *Wyeth v. Levine*, 2009).
- **Challenge:** The bill’s language is explicit in preempting state laws, which courts may respect unless it’s found to exceed congressional authority.
- **Byrd Rule in Senate:**
- If the Senate finds that Section 43201(c) violates the Byrd Rule (which limits non-budgetary provisions in reconciliation bills), it could be removed before passage. States could lobby senators to challenge the provision during Senate review.
- **Limitation:** This depends on Senate procedural dynamics and is not a direct state action.
#### 3. Non-Regulatory Approaches
States could use non-legislative or non-regulatory tools to influence AI use without violating the moratorium:
- **Voluntary Standards and Guidelines:**
- States could issue non-binding guidelines or certifications for ethical AI use, encouraging companies to adopt responsible practices without enforceable regulations.
- **Example:** A state could create a “Trusted AI” certification for companies that meet privacy or transparency standards, influencing market behavior.
- **Limitation:** Voluntary measures lack enforcement power and may have limited impact on non-compliant entities.
- **Procurement Policies:**
- States control their own purchasing decisions and could require AI vendors to meet specific standards (e.g., fairness, transparency) as a condition of state contracts. This avoids direct regulation of AI systems.
- **Example:** California could mandate that AI systems used in state agencies comply with anti-bias protocols.
- **Limitation:** This only affects AI systems procured by the state, not the broader market.
- **Public Advocacy and Coalitions:**
- States could form coalitions (e.g., through the National Governors Association) to pressure Congress to amend or repeal the moratorium, or to advocate for federal AI regulations that align with state priorities.
- **Example:** States like California and New York, with strong AI regulatory histories, could lead advocacy efforts.
- **Limitation:** This is a long-term strategy and does not immediately circumvent the moratorium.
#### 4. Refuse to Cooperate with Federal Enforcement
States could adopt a passive resistance strategy by refusing to assist federal enforcement of the moratorium:
- **Non-Cooperation:**
- States could decline to allocate resources to enforce the federal moratorium, effectively prioritizing other state laws. For example, state attorneys general could focus on enforcing criminal AI laws (which are exempt) rather than complying with federal oversight.
- **Precedent:** Some states have resisted federal mandates by refusing cooperation, such as in immigration enforcement (e.g., sanctuary cities).
- **Limitation:** The moratorium prohibits state enforcement of AI laws, not federal action, so non-cooperation may have limited impact unless paired with legal challenges.
#### 5. Wait for Legislative or Political Changes
If the bill becomes law, states could pursue longer-term strategies to undermine or outlast the moratorium:
- **Lobby for Federal Amendments:**
- States could push for future congressional amendments to narrow the moratorium’s scope or shorten its duration, especially if public backlash grows over AI-related harms.
- **Example:** States could highlight cases of AI misuse (e.g., deepfakes, biased algorithms) to build support for restoring state authority.
- **Sunset Provision:**
- The moratorium lasts 10 years from enactment. States could prepare to reintroduce AI regulations after the moratorium expires, assuming the bill passes unchanged.
- **Limitation:** This delays state action significantly and depends on future political conditions.
#### 6. Federal Funding Misinterpretation (Clarification)
The question’s earlier context about states “not accepting federal IT money” does not apply, as the moratorium is not tied to federal funding. However, to address a hypothetical:
- If the moratorium were conditioned on accepting federal funds (which it is not), states could theoretically avoid it by rejecting such funds. In practice, this is unlikely, as federal IT modernization funds ($500 million over 10 years) are separate from the moratorium’s universal application.
- **Limitation:** Rejecting federal funds would not exempt states from the moratorium under the current bill text.
### Challenges and Considerations
- **Broad Federal Preemption:** The moratorium’s language is sweeping, leaving little room for states to regulate AI outside the exceptions. Courts typically uphold explicit federal preemption clauses, reducing the likelihood of successful legal challenges.
- **Political Dynamics:** States with pro-innovation agendas (e.g., Texas, Florida) may support the moratorium, while others (e.g., California, New York) may resist, creating a fragmented response.
- **Practical Impact:** Even if states find workarounds, the moratorium could chill state-level AI regulation by creating legal uncertainty or encouraging companies to challenge state laws.
### Conclusion
States have **limited but viable options** to avoid or mitigate the prohibition on AI regulations under Section 43201(c):
1. **Most Promising Avenues:**
- Enact AI-related laws with **criminal penalties** to leverage the exemption.
- Use **non-regulatory tools** like procurement policies or voluntary standards to influence AI use indirectly.
- Pursue **legal challenges** based on federalism or preemption, though success is uncertain.
2. **Less Immediate Options:**
- Advocate for Senate rejection via the Byrd Rule or future amendments.
- Prepare for post-moratorium regulation after 10 years.
3. **Federal Funding Irrelevance:** The moratorium applies universally, so rejecting federal IT money does not provide an escape.
The original claim’s concern about states losing regulatory power is largely accurate, but states can exploit exemptions and non-regulatory strategies to retain some influence. If the bill passes the Senate unchanged, states will need to act creatively within these constraints. For real-time updates on the bill’s status or specific state responses, I can search further if requested.
BRICS and Iran
Rob on AUMF BOTG
Adam—This is really a political question rather than a question about war
crimes—but let me take a swipe. I admit that my opinion here is no
better than anyone else’s. So take it for what it is: my personal
opinion.
Let me start with what you already know. This all involves the perpetual
tug-of-war between the branches of government, which the founders built
into the Constitution. On the one hand, Article II makes the President
Commander-in-Chief, but on the other hand, Article I gives Congress the
power to declare war.
Formal declarations of war, AUMFs, and the limitations of the War Powers
Resolution all limit what the President can do. But they’re all aimed
at sustained military campaigns, not one-off operations like
Operation Midnight Hammer and all the other discrete operations that
every single president in my lifetime has done. There’s a long history
of Commanders-in-Chief taking action without Congressional permission:
Carter with Operation Eagle Claw (Iran); Reagan with Lebanon, Grenada,
and Libya; George H.W. Bush with Panama; Clinton with Somalia, Bosnia,
and I believe Kosovo; Obama with Libya (Qaddafi ouster) and Pakistan
(Osama bin Laden); and Biden in Syria and Iraq.
But, as with everything else, Trump is treated with extra disdain. I don’t
see how someone can make the case that a discrete, one-night bombing
operation on a legitimate target, carried out in a lawful manner under
LOAC, can be called impeachable. It’s definitely not a “war crime.”
But political wrangling is notoriously unpredictable, and the Democrats
are looking for any excuse to use the word “impeachment.”
My prediction: This will be debated to death until the next shiny object
comes along. And then it will fade out just like every other complaint
du jour.
353 Words on War Crimes BOTG
Adam—I heard the call. Here’s a 353-word essay on whether Operation Midnight Hammer was a war crime. Let me say right up front that when I was an Air Force JAG, I had to give “Law of Armed Conflict Briefings” to our airmen several times a year. This is a trip down Memory Lane. (Also, note that “ingenue” is pronounced _ON-juh-new_.)
No, Operation Midnight Hammer is not a war crime. Shockingly, the TikTok Ingenue Squad is wrong.
First, was the attack justified? I think so—and good luck proving otherwise. The [U.N. Charter, art. 51](https://legal.un.org/repertory/art51.shtml) expressly allows attacks done in “collective self-defence.” The U.S. takes the position that self-defensive measures can be taken anticipatorily. Here, Iran and Israel were already locked in conflict; Iran was lobbing missiles indiscriminately at Israel; Iran is a known supporter of terrorism; and Iran’s nuclear ambitions have long troubled the international community. The TikTokkers can quibble with these facts and argue that the raid was an unjustified violation of international law, but they won’t get far. The American arguments for aiding an ally and protecting American interests is credible and supported by other U.N. member states. Trump 1, TikTok 0.
Second, was the attack carried out in a lawful manner? Absolutely. This is governed by the law of armed conflict (LOAC), which derives from the Geneva Conventions and other international treaties. It has four core principles:
1. Distinction: Whenever possible, distinguish between civilians and combatants, and civilian assets versus military assets. Here, Iran itself reports no civilian casualties. Trump 2, TikTok 0.
2. Proportionality: Achieve military objectives with minimal loss of life and property. No scorched-earth unless demonstrably necessary. Here, my beloved Air Force excelled—pinpoint accuracy, damage limited to very specific military targets. Trump 3, TikTok 0.
3. Necessity: Limit damage to what’s necessary to achieve a valid military objective—do not exceed that. Here again, our military exercised incredible restraint, sticking to targets that were necessary to the objective and limiting the damage. Trump 4, TikTok 0.
4. Needless suffering: No chemical weapons, no fragmenting rounds—hit your targets with as little human suffering as possible. Here, there were no civilian casualties, no contamination, no homes blown up, etc. Trump 5, TikTok 0.
In conclusion, the United States picked a legitimate target and attacked it well within LOAC’s limits. Every step of the way, Judge Advocates were consulted to ensure compliance. Moral of the story: Don’t listen to unhinged, uninformed TikTok TDI cases. Also, ban The View.
