AI Policy in the Trump Administration and Congress after the 2024 Elections
This analysis is in response to breaking news and it will be updated. Please contact pr@rstreet.org to speak with the author.
Tuesday’s elections signaled a major shift in American politics that will have consequences for many sectors and issues, including technology policy. The return of Donald J. Trump to the White House will have some particularly significant ramifications for artificial intelligence (AI) governance. AI policy priorities also will likely change in Congress and the states in the wake of the election. The focus now will be on countering China on emerging technology while pulling back on the regulatory excesses of the Biden administration on AI policy.
The Dismal AI Narrative from the Biden Administration and the States
Interest in AI policy has exploded since Trump left the White House. The Biden administration issued a string of major reports, agency memorandums, and executive decrees focused on significantly expanding the government oversight of AI in various ways. Instead of waiting for Congress to act on AI matters, the Biden administration essentially pursued indirect AI regulation through its “Blueprint for an AI Bill of Rights,” a 110+ page executive order (EO), and a litany of new agency directives premised on combating vague notions of “algorithmic fairness” or hypothetical AI harms. These Biden administration policy pronouncements spoke of AI in dark terms, claiming that algorithmic systems are “unsafe, ineffective, or biased” and “deeply harmful.”
State-level interest in AI regulation also expanded rapidly over the past four years. Almost 750 federal and state bills have been introduced across America, with dozens becoming law. Colorado enacted the first comprehensive AI measure in May, and other states including Connecticut and Texas have considered similar policies aimed at preemptively regulating AI systems to address perceived “algorithmic discrimination.” California considered a different sort of major AI regulation focused on so-called “existential risk” scenarios that view AI as a “Terminator”-like catastrophic threat to humanity. The bill would have regulated the underlying nature and power of large AI “frontier” models. While it passed the California Legislature, Gov. Gavin Newsom vetoed the measure in late September. Many other AI-related state bills address more targeted concerns around online safety, elections, copyright, and “deepfakes,” focusing on AI as a danger rather than an opportunity. This swelling state-by-state activity puts added pressure on Congress and the new administration to create a national AI policy framework before a confusing and costly patchwork of state-level policies undermines AI innovation and speech and upends the highly successful national policy model crafted in the 1990s for the internet, digital commerce, and online speech.
AI Policy in the Previous Trump Administration
The previous Trump administration took some major actions on AI policy late in Trump’s presidency. To help promote AI capabilities, Trump built on earlier Obama administration efforts to expand federal research and development efforts and promote the use of AI within government. Toward that end, Trump signed EO 13859, “Maintaining American Leadership in Artificial Intelligence,” in February 2019. The EO created the American AI Initiative to “focus the resources of the Federal government to develop AI in order to increase our Nation’s prosperity, enhance our national and economic security, and improve quality of life for the American people.”
The Trump EO also targeted regulatory issues and sought to “reduce barriers to the use of AI technologies in order to promote their innovative application while protecting civil liberties, privacy, American values, and United States economic and national security.” Importantly, under the Trump EO, the U.S. Office of Management and Budget (OMB) was instructed to work with other executive offices to establish guidelines for federal agencies to follow when formulating AI policy. This reflected the administration’s stated desire to “limit regulatory overreach” and “promote a light-touch approach,” as Trump’s chief technology officer argued in an essay.
The EO led to an important November 2020 OMB memorandum to federal departments, “Guidance for Regulation of Artificial Intelligence Applications.” Following the Trump administration’s light-touch approach to AI policy, the OMB AI guidance directed agencies to “consider ways to reduce barriers to the development and adoption of AI technologies” and avoid regulatory actions “that needlessly hamper AI innovation and growth.” Specifically, the guidance stated, “[a]gencies must avoid a precautionary approach that holds AI systems to such an impossibly high standard that society cannot enjoy their benefits,” stressing that agencies should consider how to boost AI innovation and growth “through forbearing from new regulations” while also identifying “inconsistent, burdensome, and duplicative” laws promulgated by states and localities. The OMB guidance also encouraged agency heads to consider how “non-regulatory approaches to AI”—such as sector-specific policy guidance, voluntary frameworks and standards, and pilot programs and experiments—could better address concerns.
These pronouncements and actions stand in stark contrast to the tone and recommendations found in Biden administration policy documents. For that reason, the Trump team added to the Republican party platform this summer a promise to “repeal Joe Biden’s dangerous Executive Order that hinders AI Innovation, and imposes Radical Leftwing ideas on the development of this technology,” thus setting the stage for immediate action on AI when Trump returns to the White House.
Likely AI Trends in the New Trump Administration
Trump’s past actions and statements on AI policy, along with statements on other economic and national security policy priorities, foreshadow at least four key developments to watch for when he returns to office.
First, as reported, the Trump administration will repeal and replace the Biden AI EO and (consistent with Trump’s 2020 OMB guidance and recent Supreme Court decisions) will likely place new constraints on many agency AI regulatory actions. In doing so, however, the administration will likely retain some elements of the Biden EO, including cybersecurity guidelines, efforts to encourage agencies to use AI to improve the delivery of certain government services (and drive down costs in the process), and certain national security-related recommendations that flowed from it.
Second, there will likely be an even stronger focus on how to leverage AI as a geopolitical technological advantage over China to see which nation will become the most tech-enabled state in what some refer to as a growing “AI Cold War.” Trump’s GOP platform stressed the need to “secure strategic independence from China” as part of an expanding desire to “decouple” from the communist state on trade and technology. An earlier 2020 Trump White House report also argued that America’s “market-oriented approach will allow us to prevail against state-directed models that produce waste and disincentivize innovation” like those in China and Russia. This new approach also could entail significant pushback against efforts by the European Union and other countries to have the United States join them in advancing more international AI governance efforts, although the chances of this result are less clear.
Third, there will likely be a major nexus between AI policy and energy policy priorities, with Trump looking to capitalize on his party platform’s promise to boost “reliable and abundant low cost energy” options, which are particularly important to meet AI’s growing energy demands. We anticipate the use of AI-related priorities to advance permitting reforms and regulatory relaxation of various energy and environmental restrictions to ensure the development of more abundant energy options—especially nuclear power.
Finally, we should expect plenty of general pushback on so-called “woke AI” concerns that Trump and other conservatives have increasingly challenged in recent years. The GOP platform said that it “will stop woke and weaponized government,” and the House Select Subcommittee on the Weaponization of the Federal Government has already held hearings related to concerns about “AI-powered censorship and propaganda tools.” Although well-intended, such moves threaten to further politicize AI policy and drag it into the culture wars.
Possible AI Tensions in the Trump Administration
There will also be some interesting AI policy tensions within the new Trump administration, however, as it remains unclear how Trump and other conservatives will address some key concerns. Tension could potentially develop around the following issues:
1. Open-source AI: Friend or foe? Some conservatives look at open-source AI as a useful counterweight to the power of “Big Tech.” But national security hawks have long viewed open-source systems with suspicion and worry about potentially dangerous consequences, such as giving foreign adversaries access to powerful capabilities. This tension could play out in the growing fight over high-tech export controls, in which expansive new restrictions could negatively affect open-source capabilities.
2. Algorithmic speech: Abolish 230 or defend online speech? Trump and many other conservatives have railed against social media platforms and called for new regulations or the elimination of Section 230, an important 1996 law that protected online intermediaries from liability for the content and communications posted by others on their networks. Section 230 helped boost online communications and digital commerce, but some policymakers think the law is too broad and want it reformed or repealed because they think social media algorithms are biased against conservatives. Yet, now that Elon Musk owns X.com and Trump has benefited from his own social media site (Truth Social) and other digital outreach efforts, it is less clear how they will approach the question of algorithmic regulation or liability, especially for generative AI systems.
3. Preemption: National framework or “states’ rights”? As noted, state and local governments are aggressively pushing AI regulations. At the federal level, most policymakers have said little about how to approach this development. In particular, it remains unclear whether the new Trump team and Republicans in Congress will push for a comprehensive national AI policy framework that preempts much of this state-level regulation or defer to “states’ rights” instead, allowing a patchwork of policies to develop.
4. Industrial policy: Let private sector lead or “CHIPS Part 2”? While AI regulatory initiatives remain contentious, various high-tech industrial policy efforts have enjoyed broad-based support recently, culminating in the passage of the bipartisan CHIPS and Science Act of 2022, which included billions of dollars in subsidies for semiconductors and other high-tech priorities. Calls for a “CHIPS 2” have predictably followed. But the law is riddled with administrative inefficiencies, as the financial support promised years ago still has not materialized. The appetite for spending billions more on risky industrial policy gambits likely will be even lower in a more budget-conscious Congress. Fortunately, private-sector AI investment has reached historic levels without government intervention, which could take pressure off the new administration or Congress to spend more public money on AI initiatives.
It remains unclear how some of these tensions will be worked out by the new administration and Congress, but much of the final outcome will come down to a question of which agencies and congressional committees take the lead on AI policy. During the previous Trump administration, most AI policy was formulated by the White House Office of Science and Technology Policy. During the Biden years, however, a great deal of critical AI decision-making was delegated to the National Institute of Standards and Technology (NIST) and the National Telecommunications and Information Administration within the U.S. Department of Commerce. The Biden administration also independently created the new AI Safety Institute (AISI) within NIST.
The AISI is currently the focus of a heated debate in Congress, which is considering lame-duck session legislation that could formally bless its creation. But it remains unclear whether the Trump administration and Republicans in Congress will consent to the creation of a new AI bureaucracy devised by Biden or shift the locus of decision-making back to the White House. It also remains uncertain how much of a voice the U.S. Department of Defense will have in shaping AI policy in the new administration—a choice that could have profound bearing on the resolution of issues mentioned above, especially open-source regulation.
What is certain, though, is that Trump’s second term “promises a new era in federal AI policy” and that we will witness efforts by the incoming administration and Congress to pull back some of the amorphous new AI powers endorsed by the Biden administration. But what happens from that point on is more difficult to predict.