Cruz Calls Out Potentially Illegal Foreign Influence on U.S. AI Policy

Letter urges DOJ probe, details how Europeans are seeking to hamstring US AI development

WASHINGTON, D.C. - U.S. Senate Commerce Committee Ranking Member Ted Cruz (R-Texas) recently sent a letter to Attorney General Merrick Garland requesting the Department of Justice (DOJ) investigate whether foreign organizations had failed to register under the Foreign Agents Registration Act (FARA) pursuant to efforts to influence U.S. AI policy.

Sen. Cruz specifically points to the actions of the U.K.-based non-profit Centre for the Governance of Artificial Intelligence (Centre), which recently co-hosted an AI policy conference in San Francisco, lobbied federal lawmakers and bureaucrats about U.S. AI policies, submitted comments on various federal agency solicitations, and spoke before the U.S. Senate AI Insight Forum. Foreign groups engaged in such activities must generally register as a foreign agent under FARA, but it does not appear the Centre has done so.

Concerns about foreign involvement in shaping U.S. AI policy stem from European governments' extreme and proscriptive approach to regulating new and emerging technologies. The letter highlights how copying the Europeans' regulatory approach would hurt Texas's emerging tech industry and undermine U.S. innovation as American firms work to outpace China on AI technologies.

Additionally, the letter emphasizes how European governments have weaponized their laws to go after U.S. tech companies, like Elon Musk's X, for not censoring alleged misinformation. The Biden-Harris administration has already imposed similar censorship-enabling restrictions on AI and tech companies. The letter requests any information related to foreign entities' direct or indirect engagement in political activities to impact U.S. AI policy.

In?his letter to DOJ, Sen. Cruz wrote:?

[…] The administration is coordinating with the UK and EU on AI initiatives despite having been given no such explicit direction from Congress. In April, the U.S. signed a Memorandum of Understanding (MOU) with the UK to align their approaches to evaluating AI models, systems, and agents and in September, the administration entered into the first global treaty on AI with the EU and ten other non-member states. (Thus far, the administration has failed to provide the MOU to Congress or submit this treaty to the Senate for ratification.) Moreover, Europeans are crossing the pond to "reach out to key public and private stakeholders" about their efforts to control AI development. The EU has opened a digital technology field office in San Francisco to "further reinforce the EU's work" and the UK is opening a San Franscico office for its AI Safety Institute.

Foreign organizations, like the Centre, are also influencing U.S. AI policy, but not complying with U.S. law. FARA requires that "persons who engage in specified activities within the United States on behalf of a foreign principal register with and disclose those activities to the Department of Justice." Pursuant to this requirement, organizations that "act[] as an agent . . . at the order, request, or under the direction or control, of a foreign principal" or "whose activities are directly or indirectly . . . directed, controlled, financed, or subsidized in whole or in major part by a foreign principal" and "engage[] within the United States in political activities for or in the interests of such foreign principal" must register as representatives of that foreign principal. "Political activities" include "any activity that the person engaging in believes will, or that the person intends to, in any way influence any agency or official of the Government of the United States or any section of the public within the United States with reference to formulating, adopting, or changing the domestic or foreign policies of the United States."??

The Centre qualifies as a "principal of a foreign agent" under FARA. The Centre purports to address risks from AI, but seemingly focuses on left-wing social policy. It claims AI "risks" include "unemployment" and "inequality" as a predicate to more government regulation. The Centre also develops material to assist governments in how they can "use their regulatory toolboxes to ensure that AI developers and users behave responsibly" (ostensibly in a manner to ameliorate the "inequality" risk) and to further the role that "international coordination [can] play in reducing risks from AI." And it has "research collaborations" with the UK AI Safety Institute. The Centre is organized and primarily located in the UK and its U.S. branch is merely a shell subsidiary with a "virtual office." Most of its coporate officers and board of directors are foreign citizens. Moreover, its leadership is based in the UK. For example, Ben Garfinkel, the Centre's director who "set[s] the direction of the organisation, mak[es] key decisions, and oversee[s] its research" is a UK resident.?

The Centre also engages in the types of "political activities" that FARA covers. As noted, this week the Centre is co-hosting an invitation-only conference on "frontier AI safety frameworks" with the UK AI Safety Institute in San Francisco; it is set to begin the same day the first meeting of the International Network of AI Safety Institutes-which include, U.S. government officials-concludes in the city. The Centre's event will convene representatives from foreign governments and AI companies to discuss the design and implementation of a new global regulatory framework meant to "protect" the public from the "severe risks" posed by AI innovation. In addition, the Centre has provided public comments on the administration's requests for information regarding national AI policies and spoken before the U.S. Senate AI Insight Forum, and Centre staff have personally lobbied policymakers in Washington, D.C. ?Moreover, the Centre seems to have helped shape the "NIST AI 800-1, Managing Misuse Risk for Dual-Use Frontier Models" document-which is expected to act as a regulatory framework for companies to follow when designing AI models and cites the Centre multiple times. The Centre is also a member of the US AI Safety Institute's "consortium dedicated to AI safety," which is "developing guidelines" for AI "safety and security." The Centre therefore should register with DOJ under FARA.?

In conclusion, Sen. Cruz warns that allowing foreign nations to dictate our AI policy poses significant risks to the United States and would set us behind China in the race to lead AI innovation:

The Biden-Harris administration's efforts to tie this country's AI policies to foreign countries' agendas will have real consequences. They will set the U.S. behind China in the race to lead AI innovation. Rather than putting onerous, unnecessary regulations on AI development, China is constructing "Chinese-built AI ecosystems" that could pose risks to the United States. We should be doing everything possible to unleash competition with China, not putting up artificial roadblocks. Moreover, imposition of these foreign AI policies on American firms will pave the path for government regulation of speech and information sharing. The EU has already used newly minted technology regulations to go after X and Elon Musk for not censoring alleged "misinformation." If the radical left succeeds in carrying over the EU's AI policies, they will carry over its censorship capabilities as well-in direct violation of the First Amendment. While the Biden-Harris administration may not want to inform the American people of just how closely they are collaborating with foreign governments, at the very least, the American people have a right to know what foreign actors are trying to impose these radical regulations on American companies.?

Read the full text of this letter HERE.

###

https://www.commerce.senate.gov/2024/12/cruz-calls-out-potentially-illegal-foreign-influence-on-u-s-ai-policy