President-elect Donald Trump’s campaign and post-election transition have given several strong indications of how the new administration is likely to approach artificial intelligence (AI) policy during his second term, which we describe below. This Update is part of an ongoing Perkins Coie series about what to expect from the second Trump administration, which has included our prior Updates on the potential impact on the Federal Communications Commission, Federal Trade Commission, immigration, and retailers.
Trump’s Views on Artificial Intelligence
During the recent campaign, in a podcast interview with Logan Paul, President-elect Trump remarked that AI is a “superpower” and “alarming,” particularly with the proliferation of deepfakes. He also asserted a belief that AI proliferation was inevitable and that the United States needed to compete with China in the space. However, not all of President Trump’s comments towards AI were negative. He ended the AI-focused portion of the interview by noting the possibility of using AI for “good” and that he has used speeches edited by AI during the campaign.
BROWNSVILLE, TEXAS - NOVEMBER 19: U.S. President-elect Donald Trump and Elon Musk watch the launch of the sixth test flight of the SpaceX Starship rocket on November 19, 2024 in Brownsville, Texas. SpaceX’s billionaire owner, Elon Musk, a Trump confidante, has been tapped to lead the new Department of Government Efficiency alongside former presidential candidate Vivek Ramaswamy. (Photo by Brandon Bell/Getty Images)
Under his first administration, President Trump took a fairly laissez-faire approach to AI regulation, releasing two executive orders on AI in 2019 and in 2020. In contrast, the Biden administration has focused on the development of “responsible AI.” In 2023, President Biden issued an AI executive order that established federal guidance and policies around government purchasing and use of AI systems, as well as national security-related AI policy (Biden AI EO).
In addition, President-elect Trump has promised to revoke the Biden AI EO. He further promised to “ban the use of AI to censor speech” on day one of his administration. The Republican National Committee formally adopted this position in its 2024 platform, stating that the Biden AI EO “hinders innovation and imposes Radical Leftwing Ideas,” and advocated for the adoption of AI policy “rooted in Free Speech and Human Flourishing [sic].” The Biden AI EO requires the National Institute of Standards and Technology (NIST) to establish guidance to ensure AI models are free of biases that could discriminate against certain groups based on factors like race or gender. This role delegated to NIST appears to be what the Trump campaign was referring to as imposing "radical left-wing ideas."
The Views of Trump’s Transition Team
President-elect Trump’s transition team has selected Michael Kratsios and Gail Slater as co-chairs of tech policy during the transition.
Kratsios previously served as chief technology officer within the White House Office of Science and Tech Policy and later as acting undersecretary of defense for research and engineering during Trump's first term. He was one of the authors of Trump’s 2020 AI executive order, in which he called for increased investments in AI across many industries and for agencies to “develop and maintain approaches for the safe and trustworthy creation and adoption of new AI technologies.”
During the Biden administration, Kratsios has been serving as a general advisor at Scale AI, an AI startup that supports the defense sector. His return indicates the Trump administration holds a continued interest in the intersection of defense and AI. In 2023, Kratsios advocated for federal agencies to “pursue a use case and sector-specific, risk-based approach rooted in high quality testing and evaluation.”
Gail Slater previously served as an economic policy advisor to Senator JD Vance, who generally advocated for less regulation of the AI industry and increased oversight of large technology companies. Prior to her role in Vance’s office, Slater worked at Roku, Fox Corp., and the now-defunct trade group The Internet Association (whose members included many large tech platform companies). On December 4, 2024, Slater was reportedly selected as the new chief of the U.S. Department of Justice’s Antitrust Division. The elevation of Kratsios and Slater may suggest we should expect a light regulatory touch for AI under the new Trump administration and a strong interest in national security-related AI systems.
What It Means To Revoke the Biden AI EO
The Biden AI EO created an overarching executive branch policy for AI that binds federal agency AI purchases and uses. Revoking the order, in whole or part, may result in the termination of these policies, along with their implementation and funding. For example:
AISI. The Biden AI EO established the U.S. AI Safety Institute (AISI) and gave it some of the authority previously held by the Office of Science and Technology Policy.
Revoking without legislation to preserve the institute would likely mean that the AISI would be dissolved, and its powers would return to the Office of Science and Technology Policy.
The AISI has bipartisan support. Both the Senate and House have pending bills to either establish the institute or preserve the functions of the institute under a new name. Such bills are unlikely to pass before the beginning of the new administration but could be passed during the Trump administration so long as both legislative houses remain controlled by the same party.
OMB. The Office of Management and Budget (OMB) released binding guidance for agency procurement and use of AI pursuant to the Biden AI EO.
The OMB memo imposed significant requirements for federal agencies regarding the acquisition and management of AI systems, especially systems that are deemed to be rights- or safety-impacting. These requirements were expected to increase the compliance costs of potential government contractors who sell AI systems and products to federal agencies.
While revoking the Biden AI EO would not automatically rescind the guidance, it is unclear if the guidance will remain binding without the legal support of the order. Further, it is likely the Trump administration will explicitly rescind some, or even most, of the OMB requirements, such as the requirement to consider climate impacts.
Chief AI officers. The Biden AI EO called for the appointment of chief AI officers at federal agencies, some of whom have already been appointed.
While it is not uncommon for senior federal officials to be replaced by a new administration, repeal of the Biden AI EO might eliminate the positions entirely.
Regulation of “woke” AI. Given President-elect Trump and other Republicans’ statements against the so-called “woke” efforts of the Biden administration, the new administration will likely cease executive efforts to mitigate the risks of discrimination, bias, and climate change.
National Security and the Role of China
Under President Biden, export controls to China have been repeatedly tightened on AI semiconductor chips. Given that President-elect Trump has advocated taking a hard line against China in matters of trade (e.g., proposing significant tariffs against products manufactured in China), we can expect the next administration to build upon current controls and seek significant protectionist measures to deter the sale of sophisticated AI technologies to China.
Such restrictions would likely have bipartisan support. On November 19th, 2024, the U.S.-China Economic and Security Review commission issued a report to Congress recommending the establishment of a “Manhattan Project-like program dedicated to racing to and acquiring” artificial general intelligence capabilities. Specifically, the report recommended that Congress should:
Provide the executive branch with broad contracting authority and funding.
Direct the secretary of defense to provide a “Defense Priorities and Allocations System ‘DX Rating’” to AI items to ensure AI is a "national priority.”
Similarly, the Biden administration recently released a national security memo on AI, which mirrored parts of the European Union’s AI Act in requiring risk assessments. While the Trump administration will likely scale back parts of the framework that require assessment of risks to human liberties, other core elements of the framework could remain in place until the Trump administration releases its own executive orders.
The AI Czar
President-elect Trump is reportedly considering naming an official AI czar to steer federal policy on the topic. According to Axios, Elon Musk and Vivek Ramaswamy will have a significant role in the selection process. No potential candidates have been publicly identified at this time.
This will not be the first time America has had an AI czar. President Biden appointed the first AI czar, Lynne Parker, to her role as director of the National Artificial Intelligence Office in 2021. In that role, she focused on establishing safeguards around AI use and finding nonpartisan solutions to “reduce the potential harms of AI.”
The new czar will likely be tasked with coordinating with federal agencies and outside stakeholders to ensure consistent guidance regarding AI use in the federal government. Presumably, they will seek input from Musk, Ramaswamy, and the co-chairs of tech policy in crafting the Trump administration's stance on AI.
A Focus on Innovation and “Little Tech”
Rather than focusing as much as the Biden administration did on risks and potential harms of AI, the Trump administration is likely to focus more on promoting and facilitating AI innovation by U.S. industry with fewer restraints. This shift may manifest in the form of deregulatory measures and significant reductions in key federal agencies. It will also be seen through the new administration no longer supporting “responsible AI” initiatives as discussed above under the Biden AI EO.
Relatedly, some tech industry advocates are asking the incoming administration to consider a “little tech” agenda—i.e., to consider what policies will best enable a thriving ecosystem of a broad range and significant number of tech startups, particularly in areas like AI, rather than ceding the space of AI innovation to a handful of large tech companies.
Federal AI Legislation
Congress failed to pass comprehensive AI legislation during the Biden administration. Now that both houses of Congress are controlled, albeit narrowly, by the same party as the White House, there is a window of opportunity. It is unclear, however, whether the Trump administration will support a federal legislative response to the EU’s AI Act or new federal laws to mitigate the risks of sexually explicit and election deepfakes. For example, during the presidential campaign, Elon Musk posted a deepfake video on the social media platform X in which Vice President Kamala Harris appeared to say things she did not actually say. In addition, because California’s governor vetoed a comprehensive AI bill, SB 1047, there does not appear to be an urgent need to enact a federal law to preempt state AI laws that might be inconsistent with the Trump administration’s AI policy, but this will be an area to monitor if several states enact sweeping AI legislation.
The Elon Musk Factor
The potential influence of Elon Musk in the new administration cannot be underestimated given his close relationship to the president-elect and his expected role in the administration. Before Musk’s appointment to the new Department of Governmental Efficiency, the Americans for Responsible Innovation group launched a petition calling for President Trump to appoint Musk as a special advisor on AI.
In the past, Musk has been a proponent of responsible AI and AI regulation and has cautioned against its potential risks. However, Musk is also a fierce opponent of what he claims are “woke” AI systems. His own startup, xAI, recently released Grok, a proclaimed “antiwoke” AI chatbot with fewer guardrails than other chatbots. We will have to see how these seemingly competing views play out in the coming years.
Takeaways
With President-elect Donald Trump’s victory and the Republicans gaining control of both houses of the legislative branch, significant policy shifts in AI regulation are expected. The new administration is likely to substantively revise, if not fully replace the Biden AI EO—particularly its provisions related to preventing discrimination, bias, and climate change. Instead, the focus will shift towards national security and economic competition, with an emphasis on promoting innovation and reducing regulatory barriers.
Stakeholders should prepare for significant changes in AI policy that emphasize America’s leadership in AI innovation, defense, and national security. These shifts reflect the incoming administration’s long-standing preference for deregulation and also signal a commitment to follow through on campaign promises to fight “wokeness,” including with federal AI policy. The administration's stance on AI is expected to foster private investment and reduce regulatory barriers, aligning with President-elect Trump’s broader economic and policy objectives.
[View source.]
Kommentare