I don’t actually know Adam Thierer, however I did cross paths with him a pair instances a few years in the past in DC. Like so lots of the tightly-wound libertarian triumphalists which might be interested in the anti-copyright crowd, my first impression of him was that he ought to have auditioned for The Accountant. Absolutely he’d a minimum of have gotten a desk learn. I in all probability thought that as a result of he appeared very uneasy—which I attributed to a fear about getting caught when actually it extra seemingly was simply that he was a lot smarter than me. However solely one in every of us knew it.
Fairly naturally it appears, Mr. Thierer has lately discovered his strategy to the R Avenue coverage laundry that’s as near Google as one is to 1.01. Readers will recall that R Avenue was one of many braying hounds who went after then-Mississippi Lawyer Common Jim Hood in a hateful show again in 2016. This was earlier than Mr. T arrived, however was when Common Hood went after Google for violating the phrases of Google’s nonprosecution settlement on selling the sale of medicine on-line and its $500,000,000 advantageous paid with the stockholders cash (for which Google was additionally sued by shareholders).
That’s proper—R Avenue went after Hood, not Google. (We lined that saga fairly extensively on the time. Full disclosure, Common Hood invited me to take part on his panel on the Nationwide Affiliation of Attorneys Common the place I introduced on how Google promotes model sponsored piracy and pro-terror movies and pimp recruitment on YouTube. That will be the identical NAAG that despatched a letter opposing Mr. T’s secure harbor signed by 40 attorneys basic–the identical variety of AGs who signed onto an amicus transient defending Jim Hood towards Google. Simply sayin’.)
So it got here as no shock that Mr. Thierer appears to be the prime mover within the cowl story for the loathsome AI secure harbor moratorium now making its method by way of Congress as a part of the One Large Lovely Invoice Act. How do we all know this? Mr. Thierer has been one of the crucial prolific voices calling for a federal “studying interval” moratorium—a euphemism for legislative paralysis. In his writings for R Avenue and retailers like Lawfare, Mr. Thierer insists that solely the federal authorities ought to regulate “basic function” AI. Might that be as a result of it’s simply too laborious for Google to regulate all 50 state legislatures they usually already personal the feds?
Might or not it’s…the Pink Menace?
His main contributions on this subject embody: “Getting AI Coverage Proper By way of a Studying Interval Moratorium” (R Avenue, Might 2024): https://www.rstreet.org/commentary/getting-ai-policy-right-through-a-learning-period-moratorium/
He treats us to this fallacious opening paragraph, thus demonstrating that within the coverage laundry, writing is supposed to be quoted, not learn critically. He opens with the Pink Menace Concept however will get to the climax just a little too rapidly for his personal good:
Whereas some synthetic intelligence (AI) critics need to pause AI growth, the pause most wanted at the moment is on overzealous regulatory proposals that might kneecap America’s lead in computational science and algorithmic applied sciences. With over 700 federal and state AI legislative proposals threatening to drown AI innovators in a tsunami of purple tape, Congress ought to take into account adopting a “studying interval” moratorium that may restrict burdensome new federal AI mandates in addition to the looming patchwork of inconsistent state and native legal guidelines.
The time to take action is now, with the race for AI supremacy towards China intensifying and different nations investing closely to counter america.
In order that’s bullshit for starters. The concept that Silicon Valley monopolist patriots are going to go after the PRC is type of laughable, don’t you assume?
To not point out that if you happen to test Mr. Thierer’s personal supply, he fails to say that of 107 federal AI payments launched in 2024, precisely zero had been enacted. Good job R Avenue!
Of the 635 state payments launched in the identical interval, 99 had been enacted. Which exhibits you the issue—not even Google should buy each state legislature! We will’t have that, now, can we?
However see what he did there? Mr. T jumped from the “over 700” payments in Mr. Thierer’s neighborhood however that got here all the way down to 99 in actuality which he didn’t point out. Fairly an undisclosed soar. That type of shammery by no means stopped a monopolist on a mission and it’s not stopping Google it its quest for doing regardless of the heck it needs to do along with your information, their new oil.
However wait—there’s extra. By some means we soar from 700 payments to 1,000 payments in a 12 months. In his Lawfare post in Might 2025 (with Kevin Frazier), Mr. T tells us that it’s not time to for Congress to get critical about AI regulation, no, no, no—it’s time for Congress to get critical about PREEMPTION. And proper there his argument carries the seeds of its personal destruction due to that…whatchamacallit…Structure factor as others have additionally famous. You possibly can’t have one with out the opposite—in any other case there’s nothing to do the preempting.
Oh No! Not the Patchwork!
As a result of Congress has not acted in AI regulation, it doesn’t imply that there’s no harms from AI, it extra seemingly simply implies that Google et al have a extra environment friendly lobbying machine on the federal stage which I think everyone knows they do. So it must be fairly pure that the a number of States would reply to their residents with AI regulation; you already know, that entire “laboratories of democracy” factor in response to Mr. Justice Brandeis.
So both Mr. T didn’t grasp the idea of correct preemption, or possibly what he actually means is he needs us to need no regulation in any respect. As a result of even when Congress has not acted there’s no car for preemption there’s nonetheless the entire commandeering concern which we’ll get to.
Mr. T tells us that what we actually need to keep away from is a “patchwork” of state legal guidelines, so let’s check out the overall classes of those 99 state legal guidelines in response to the Nationwide Council of State Legislatures—which is strongly against what Mr. T calls a “studying interval” and what I name a loathsome secure harbor moratorium snuck into should cross laws like a Christmas bow on a pig.
So to recap, in 2024, U.S. state legislatures enacted 99 legal guidelines associated to synthetic intelligence, addressing numerous points of AI deployment and governance. These legal guidelines might be categorized into a number of key areas:
Client Safety and Transparency
States targeted on safeguarding shoppers from potential harms posed by AI programs, particularly in sectors like schooling, employment, finance, healthcare, and housing. Laws on this class usually requires corporations to reveal AI utilization, guarantee equity, and supply avenues for recourse.
Algorithmic Discrimination and Equity
A number of legal guidelines purpose to forestall AI programs from perpetuating or exacerbating biases. These embody necessities for affect assessments, bias audits, and the implementation of measures to mitigate discriminatory outcomes.
Deepfakes and Artificial Media
In response to issues in regards to the misuse of AI-generated content material, states enacted legal guidelines just like the ELVIS Act in Tennessee concentrating on the creation and distribution of deepfakes, notably those who might affect elections or contain non-consensual specific materials.
Authorities Use of AI
Laws on this space addresses how state businesses and departments can deploy AI applied sciences. Legal guidelines usually mandate transparency, accountability, and the institution of requirements for moral AI use in public companies.
AI in Healthcare
Recognizing the rising function of AI in healthcare, some states handed legal guidelines specializing in affected person rights, information privateness, and the validation of AI instruments utilized in medical settings.
Employment and Labor
To guard employees, sure legal guidelines regulate using AI in hiring, monitoring, and efficiency evaluations, making certain that automated programs don’t unfairly drawback staff.
Training and Workforce Improvement
States additionally addressed the combination of AI in academic settings and the necessity for workforce coaching applications to arrange people for AI-related job markets.
These legislative efforts mirror a complete strategy by states to handle the multifaceted implications of AI applied sciences throughout numerous sectors—they usually had been all handed on the identical time that AI was attracting the assist of the World Financial Discussion board and its bubble riders in addition to large funding within the sector by folks like SoftBank and, after all, Google itself.
Mr. Thierer argues that state-level AI payments have created a patchwork of guidelines that stifle innovation, you already know, type of like copyright stifles piracy. However in apply, Mr. T is advocating a federal freeze on state energy to behave on behalf of their residents. That is acquainted floor for R Avenue which made an analogous argument when Jim Hood tried to implement the legislation towards Google. However this time it’s on a vastly bigger scale of soul crushing exploitation.
Constitutionally Suspect: Federalism Below Siege
So again to that entire Structure thingy. Mr. T’s studying interval qua secure harbor qua AI moratorium violates two core constitutional doctrines:
Commerce Clause Overreach: It treats all AI exercise as interstate commerce, even native or non-commercial use. I believe you’ll discover that the Supreme Court docket has rejected such pretextual use of the Commerce Clause, notably to compel commerce corresponding to in NFIB v. Sebelius to call one.
Anti-Commandeering Doctrine: In Murphy v. NCAA, the Supreme Court docket dominated that Congress can not forestall states from legislating. However that’s precisely what the AI moratorium does: it freezes all state lawmaking with out offering a federal alternative. Mr. T got here up with a Constitutional twofer—stopping states from legislating within the absence of federal laws however calling it preemption when there’s no preemption car of federal legal guidelines to battle with state legal guidelines. Man, you speak about battle averse!
So sadly, I’ve to name bullshit on this entire angle. This moratorium shouldn’t be about considerate AI coverage—it’s about immunizing tech platforms like Google from future scrutiny. By banning states from performing, it clears a path for AI companies to function beneath federal inaction, protected by Part 230, arbitration clauses of their phrases of service powered by unilateral amendments to privateness insurance policies permitting them to do no matter they need like prepare on gmail, and now, moratorium-compelled silence.
I don’t need to denigrate the Trojan Horse by comparability, however let’s simply say Odysseus he ain’t.
The Actual Patchwork Is Company Energy
Mr. Thierer warns of a ‘patchwork’ of state legal guidelines, however the true patchwork is one in every of de facto company immunity for monopolists. I believe it’s honest to say that AI platforms function in authorized grey zones created by Congress’s refusal to behave created in flip by Google’s large lobbying (and all the remainder of them, in equity, together with Elon The Rocket Catcher). To not point out a greed-driven willingness to push folks round who can’t afford to combat again after which rolling over those who can.
States have traditionally been the primary movers on tech threat—from biometric privateness to information safety. The moratorium flips that apply into the other way up world.
The folks should see the secure harbor moratorium for what it’s: a federal veto on democratic experimentation fueled by the monopolist foyer. Constructed on astroturf coverage laundry retailers masquerading as libertarian theorists and backed by the most important of Large Tech, it will lock in silence whereas the harms of AI proliferate—and belief me, each one in every of these corporations will increase the AI secure harbor as a protection in all their litigation, present and future. Possibly even in present litigation just like the multidistrict social media adolescent dependancy case since all of them in all probability use AI to additional the dependancy on the coronary heart of the case that features the same old suspects aka Meta Platforms, Inc., Instagram LLC, Snap, Inc., TikTok, Inc., ByteDance, Inc., YouTube LLC, Google LLC, and Alphabet Inc.
A federally imposed AI secure harbor promoted by monopolists shouldn’t be libertarianism — it’s company protectionism, backed by purchased and paid for federal energy, suppressing each particular person rights and aggressive markets. They’re simply discussing the value. True liberty requires duty, restitution, and the liberty of states and people to say no. Now the place have I head that earlier than? “The powers not delegated to america by the Structure, nor prohibited by it to the States, are reserved to the States respectively, or to the folks.“ These final 4 phrases are a few of the most essential within the Structure.
Congress should reject any invoice that bars states from performing within the public curiosity—particularly when that bar originates from assume tanks with longstanding ties to the very platforms in search of immunity.
