The code cops are coming for artificial intelligence (AI) and proposing an enormous latest national and international information control system for digital technologies.

The U.S. Senate Judiciary Committee held a hearing this morning on, “Oversight of A.I.: Rules for Artificial Intelligence,” through which members and witnesses sketched out a blueprint for a comprehensive regulatory regime for AI. A number of the potential elements of this latest regime that were floated at today’s hearing include:

· A latest federal regulatory bureaucracy akin to a “Food & Drug Administration for algorithms;”

· A federal operating license for absolutely anything that lawmakers consider “dangerous,” enforced by the brand new agency;

· “Dietary labels” for AI, or some type of mandated disclosure statement about how every algorithm works;

· Mandatory algorithmic audits (conducted by third parties or agencies) after product release;

· A move to harmonize US law with European Union AI rules or other global AI regulatory regimes;

· A global regulatory body of some sort to oversee global AI development, perhaps through the United Nations;

· Other amorphous prior restraints and limitations on the use of assorted algorithmic innovations.

Taken together, this represents the start of a “Mother, May I?” permission slip-based regulatory regime for computational technology that may slam the brakes on AI innovation and concurrently open the door to a digital information control regime. It’s going to decimate America’s innovation culture and put life-enriching innovation in danger by treating AI developers as guilty until proven innocent under a legal standard of “unlawfulness by default” for algorithmic applications.

It is important that america be a frontrunner in AI to make sure our continued global competitive standing and geopolitical security. We must avoid overly burdensome technology regulations that may undermine AI’s advantages to the general public and the nation as a complete because there may be a compelling public interest in ensuring that algorithmic innovations are developed and made widely available to society.

Yet, every part about today’s Senate AI hearing suggests that America is able to ignore all that and head in exactly the other way with a costly, convoluted, top-down regulatory regime for one of the essential technologies of our lifetime.

The hearing included references to the atom bomb and various dystopian scenarios, featuring an infinite array of worst-case claims about job loss, consumer manipulation, election fraud, and other “disinformation” issues. At one point, Sen. John Kennedy suggested that these discussions should begin with the belief that AI desires to kill us. The various potential advantages related to AI, machine learning, and robotics got only passing consideration by a few a members.

Thus, it is evident that “AI [is] in Washington’s crosshairs” now and so much more regulatory intervention is on the best way. The Biden administration is concurrently moving to say greater control over AI and algorithmic systems through quite a lot of agencies and efforts. Importantly, today’s session happened against the backdrop of a much wider congressional effort to demonize the web, social media, and digital technologies typically. Those attacks have softened the bottom for a strike on AI and algorithms, which could constitute a backdoor way for Congress to control tech firms and digital platforms while attempting to avoid First Amendment scrutiny. The clear implication of all that is that principally every member of Congress thinks the web sucks and that, if we needed to do all of it once again, we must always have just imposed the old broadcast media regulatory model on digital tech, complete with licenses and censorship of speech.

At today’s hearing, Sen. Chris Coons specifically drew the linkage between AI and social media, arguing that Congress, “cannot afford to be as late to regulating generative AI” as with earlier digital technologies and platforms. He and other Senators implied that the times of web freedom were numbered and that AI innovators could be delivered to heel next.

Sadly, the witnesses generally went together with most of this, and nobody bothered seriously standing up for the liberty to innovate — or document the amazing success that the U.S. has enjoyed within the digital economy. The Latest York Times noted that, while most tech congressional hearings “can best be described as antagonistic” — an understatement, to say the least — today’s AI hearing was practically a lovefest. That’s primarily since the witnesses, and particularly OpenAI CEO Sam Altman, were quick to leap on the regulatory bandwagon. Indeed, in over 30 years of covering tech policy in Washington, I cannot recall seeing any technology executive throw his entire industry under the bus quite as fast as Sam Altman did at today’s AI hearing. He sketched out a plan for wide-reaching AI control that will decimate algorithmic innovation and digital competition. Unsurprisingly, Senators were falling throughout themselves to praise him for his capitulation.

Sen. Dick Durbin referred to the session as “historic” when it comes to the best way tech firms were coming in and begging for regulation. Durbin said firms were telling him and other members, “Stop me before I innovate again!” which gave him great joy, and the one thing that mattered now he said is “how we’re going to attain this.” He then also endorsed each a national and a worldwide regulatory authority for AI, and many other regulations.

That’s a horrifying thing to be comfortable about, yet it represents the present state of pondering in Congress about digital technology matters. Technological stagnation is now welcomed or really helpful by our national leaders.

Sen. Richard Blumenthal, who chairs the committee, noted that this is able to be the primary of many coming hearings on AI policy issues, and said he looked forward to exploring algorithmic regulation on many other fronts. Sen. Josh Hawley, the Republican minority chair on the Judiciary Committee, piled on and suggested that the very best plan of action was to open the floodgates of litigation and invite trial lawyers to file lots more lawsuits to intimidate and hobble AI developers. “The specter of litigation is a strong tool,” he smiled.

One witness, Christina Montgomery of IBM, deserves some credit for pushing back a bit against a few of this insanity. She highlighted how a more targeted, risk-based approach to AI governance is already developing, and that AI innovators have taken steps to bake-in best practices by design. [Here’s a 20,000-word study on “Flexible, Pro-Innovation Governance Strategies for Artificial Intelligence” in which I document all that activity as well as all the state capacity that already exists to address AI risks.] But the general thrust of today’s session was too caught up in techno-panicky talk and extremist proposals to be concerned with such facts and more reasoned approaches to AI governance.

Perhaps the one shiny spot through the hearing was the undeniable fact that several lawmakers and witnesses understood that regulatory capture, which occurs when special interests co-opt policymakers or political bodies to further their very own ends, could turn into an actual problem with a latest AI regulatory body. Sen. Blumenthal was even willing to concede that there was an actual “monopolization danger” related to latest agency regulations that might exclude latest competition.

After all, those aren’t the one downsides of a hypothetical “FDA for algorithms.” Lawmakers could have also identified the deleterious impact of preemptive, precautionary regulation on consumer alternative and the creation of life-enriching goods and services. Over-regulation of AI may also have profound ramifications for America’s geopolitical competitiveness and security. China and other nations are hoping to not get left within the dust by America again in the following great technological revolution. It might be shocking if America shot itself within the foot as that race got underway.

If today’s hearing was any indication, nonetheless, that appears to be exactly where America is heading. Even worse than the thought of an FDA for algorithms is the proposed global regulatory authority to in some way monitor global AI innovations, possibly through the United Nations. A fast reminder: the UN recently allowed North Korea (the world’s biggest nuclear pariah state) to take over as head of the UN Conference on Disarmament, after which let Russia take the lead on the UN Security Council after they invaded Ukraine. So, how’s this latest global regulatory body for AI going to play out once countries like those plus China, Iran, and others start making demands for “harmonizing” global algorithmic speech and commerce standards?

This can be a truly terrible, destructive model for AI that should be rejected before we commit technological suicide as a nation and squander the amazing advantages that may accompany the Computational Revolution that awaits us. Now we have in some way forgotten that the best of all AI risks is shutting down AI altogether.

:

• Adam Thierer, “,” R Street Institute Blog, May 15, 2023.
• podcast: “?” Adam Thierer and Matthew Lesh discuss artificial intelligence policy on the Institute for Economic Affairs podcast, May 6, 2023.
• Adam Thierer, “,” Medium, May 4, 2023.
• Adam Thierer, “,” Medium, April 23, 2023.
• Adam Thierer, “,” R Street Institute Policy Study №283 (April 2023).
• Adam Thierer, “,” The Hill, April 16, 2023.
• Adam Thierer, Brent Orrell, & Chris Meserole, “,” AEI Ideas, April 6, 2023.
• Adam Thierer, “,” R Street Institute Policy Study №281 (March 2023).
• Adam Thierer, “?,” R Street Institute Policy Study №278 (March 2023).
• Adam Thierer, “,” R Street Blog, March 9, 2023.
• Adam Thierer, “,’” U.S. Senate Homeland Security and Governmental Affairs Committee, March 8, 2023.
• Adam Thierer, “?” Medium, February 20, 2023.
• Adam Thierer, “,” Medium, February 14, 2023.
• Adam Thierer, “,” R Street Blog, February 9, 2023.
• Adam Thierer, “,” Medium, December 2, 2022.
• Neil Chilson & Adam Thierer, “,” Regulatory Transparency Project of the Federalist Society, November 2, 2022.
• Adam Thierer, “?” Discourse, October 6, 2022.
• Adam Thierer, “,” Medium, September 13, 2022.
• Adam Thierer, “,” Medium, September 10, 2022.
• Adam Thierer, “,” Technology Liberation Front, August 1, 2022.