Overview of Decoder with Adam Schiff
In this Decoder interview, Senator Adam Schiff argues that the U.S. is badly behind on regulating AI, and that the country needs real legislation, stronger transparency rules, and possibly a dedicated AI regulator with expert authority. He connects AI policy to broader themes of corruption, corporate power, privacy, antitrust, and the future of work, while also making the case that Democrats should focus on lowering costs, protecting workers, and confronting what he sees as Trump-era self-dealing.
Main themes
AI needs real regulation, not voluntary pledges
Schiff is dismissive of the Trump administration’s non-binding AI “safety” agreement with tech CEOs, calling it essentially meaningless. He argues that:
- AI models are advancing too quickly for informal pledges to matter.
- Frontier labs don’t fully understand or control their systems.
- AI poses major national security risks, including misuse in defense, hacking, surveillance, and autonomous weapons.
- Congress needs to move from voluntary commitments to enforceable rules.
He repeatedly frames AI as a problem of both inputs and outputs:
- Inputs: companies vacuumed up copyrighted material and training data first, then worried about legality later.
- Outputs: companies now need clear rules about safety, liability, and harmful use.
Copyright, IP, and training data are central
As ranking member on Senate committees touching IP and tech, Schiff says AI regulation must start with the question of what data companies used to train their models.
Key points:
- Publishers, studios, and creators deserve transparency about whether their work was used.
- The legal fight around fair use will depend on whether companies kept records of training inputs.
- He sees major tension between AI companies and industries whose value depends on intellectual property.
- He also notes that AI can be useful in IP-related work, like speeding up patent review.
Social media regulation failed; AI can’t be handled the same way
Schiff contrasts AI with the long failure to regulate social media. His view:
- The biggest problem with social media was not just the First Amendment.
- Section 230 gave platforms broad immunity that let them avoid responsibility.
- Platforms used free-speech arguments strategically to block meaningful regulation.
- Social media companies trained algorithms for addiction and allowed widespread harms, especially to young people.
He says AI may not fit neatly into the Section 230 framework, but the lesson is clear: Congress should not repeat the same slow, ineffective response.
A new AI regulatory agency may be necessary
Schiff supports creating an agency similar to an FDA for AI, with:
- real expertise,
- real authority,
- and the ability to move faster than Congress.
He says AI evolves faster than lawmaking can keep up, so a dedicated regulator is likely necessary. But he also acknowledges major obstacles:
- staffing such an agency with enough technical talent,
- and surviving legal challenges under the current Supreme Court.
The Supreme Court makes agency regulation harder
On the post-Loper Bright landscape, Schiff says Chevron deference is gone and courts are now much more willing to second-guess agencies. His concern is that:
- any AI rules drafted too broadly will be tied up in litigation,
- corporate interests will challenge agency power,
- and the Court often sides with moneyed interests.
That means any AI law has to be drafted carefully and specifically.
Privacy, surveillance, and public backlash
People want privacy in public now
Schiff says the public is reacting against always-on surveillance tools like Flock and Axon cameras. He argues that people are increasingly uncomfortable with:
- camera networks tracking movement,
- databases that make public life searchable,
- and AI-enhanced surveillance by law enforcement or private actors.
He suggests that existing privacy law is lagging behind social expectations.
AI could erase what little privacy remains
His deeper warning is that AI can be used to:
- infer health records,
- activate microphones on devices,
- aggregate behavioral data,
- and build invasive profiles at scale.
He warns that without action, “there’ll be little privacy left.”
Antitrust, coordination, and narrow exemptions
Schiff supports only a very narrow antitrust exemption for AI companies to share information on legitimate security threats, such as:
- coordinated cyber defense,
- responding to mass model distillation,
- or sharing threat intelligence after attacks.
But he strongly opposes any exemption that would:
- enable price fixing,
- create a cartel,
- or reduce competition under the guise of safety.
Jobs, data centers, and public anger at AI
Schiff argues the political backlash to AI is not purely partisan. In his telling, concerns are emerging across the spectrum around:
- job displacement,
- data centers,
- energy and local costs,
- and the fear of runaway AI.
He says many of the AI bills being introduced are bipartisan, and that resistance is growing as communities experience the tangible effects of AI infrastructure and automation.
Trump, corruption, and impeachment
Schiff sees corruption everywhere
He begins the interview by describing the Trump administration as deeply corrupt, pointing to:
- taxpayer-funded campaign-style ads,
- exclusion of press,
- vanity projects,
- and deals that blur public and private interest.
Would Democrats impeach Trump again?
Schiff does not commit to a third impeachment. Instead, he says Democrats first need to:
- win Congress,
- focus on lowering costs,
- and investigate whether corruption is driving up costs for ordinary Americans.
He suggests impeachment would depend on the priorities at the time, but he does not rule it out.
Entertainment, mergers, and California
Schiff also talks about media consolidation and entertainment-industry layoffs, especially in California. He is worried about:
- editorial independence at merged companies,
- layoffs from big media mergers,
- and the pressure on studios to curry favor with the Trump administration.
He’s especially critical of settlements or concessions made to the president over litigation and merger approvals, calling that kind of behavior corrosive to the First Amendment and democratic norms.
Broader political realignment
Schiff believes the U.S. is in the middle of a major political realignment driven by:
- concentration of wealth,
- distrust of both parties,
- dissatisfaction with the economy,
- and the feeling that working people are being left behind.
His prescription for Democrats:
- become more clearly the party of working people,
- push bold reforms, not just restoration,
- and build a more progressive tax system that makes wealthy beneficiaries of the system pay more.
He sees AI as part of that larger conflict: a technology that could produce enormous gains, but only if the rules protect workers, consumers, privacy, and the public interest.
Key takeaways
- Schiff wants real AI legislation, not voluntary industry promises.
- He sees copyright, privacy, cybersecurity, and antitrust as intertwined AI issues.
- He believes Congress should create a specialized AI regulator with expertise and enforcement power.
- He thinks the public is moving ahead of the law on surveillance and privacy.
- He views the current political moment as a fight over corruption, wealth concentration, and who the economy serves.
Bottom line
Schiff’s core message is that AI cannot be left to self-regulate. In his view, the U.S. needs enforceable rules, a stronger regulatory state, and a political reset that puts workers, privacy, and accountability ahead of corporate and executive power.