On 23 January 2025, US President Donald Trump issued an executive order about artificial intelligence (AI) aimed at advancing the United States’ position in the AI space. The executive order calls for the development of an AI action plan and seeks to address actions under the previous administration that conflict with this order.1 Just 3 days earlier, President Trump revoked Executive Order 14110, Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence from the previous presidential administration. This revocation occurred less than 2 years after former US President Joseph Biden issued the order.2
While executive orders apply to the federal government and not necessarily private industries, they have the power to shape norms and expectations in the country, and this can affect the way in which enterprises operate. The rapid shifts in the way the US executive branch addresses AI under different administrations illustrates why organizations are better positioned if they approach emerging technologies from risk-based and ethical perspectives.
Comparing Executive Orders
Executive Order 14110 addressed several facets of AI and AI development, including:3
- Safety and security
- Responsible development and competition
- The labor market and talent development
- Civil rights
- Fraud and discrimination
- Privacy and civil liberty
The newest executive order on AI has provisions to review any action taken related to the revoked executive order for conflict with the current administration’s approach to AI. However, there may be some overlap between the previous and current administration’s AI priorities; for example, both agree on the importance of using federal land for data centers.4
The Trump administration’s Executive Order Removing Barriers to American Leadership in Artificial Intelligence is not prescriptive. The group developing the AI action plan has 180 days to present their plan to the president,5 so more detailed guidance can be expected in the future.
The Biden administration issued 2 AI-related executive orders in their final week. As of the time of this writing, the Executive Order on Advancing United States Leadership in Artificial Intelligence Infrastructure6 and the Executive Order on Strengthening and Promoting Innovation in the Nation’s Cybersecurity7 remain in effect.
Takeaways
Federal agencies must effectively navigate change due to administrative turnover. The impacts of government-directed executive orders can trickle down to private organizations as well, and the following are considerations for organizations that are not subject to executive orders.
Laws, regulations, executive orders, and political dialogue around technology should not be the basis for how enterprises develop and/or implement technology. As has been demonstrated in the United States, norms and conversations around emerging technology can shift rapidly. The newest executive order on AI indicates future guidance may come, so enterprises relying on this to guide AI use are stuck in limbo—not being able to move forward in light of the absence of priorities.
Additionally, the guardrails set forth by Executive Order 14110—which some enterprises may have already begun to consider—were revoked. It is possible that the current administration may incorporate some guidance in these areas, but it is worth noting that the government’s approach to, and priorities regarding, AI are not static.
Organizations that only look to federal guidance to shape their AI programs may find themselves continuously working toward constantly moving goalposts. In contrast, enterprises that develop AI programs and align budgets in line with their own risk tolerance, ethical standards, and enterprise objectives will be better positioned to adopt this technology in a safe, responsible way, while minimizing the impact associated with the ebbs and flows of the regulatory landscape.
Endnotes
1 The White House, Removing Barriers to American Leadership in Artificial Intelligence, USA, 23 January 2025
2 The White House, Initial Rescissions of Harmful Executive Orders and Actions, USA, 20 January 2025
3 The White House, Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, USA, 30 October 2023
4 Shivaram, D.; “Trump has Tossed a lot of Biden Policies. But There's One AI Measure He Wants to Keep,” NPR, 24 January 2025
5 The White House, Removing Barriers to American Leadership in Artificial Intelligence, USA, 23 January 2025
6 The White House, Executive Order on Advancing United States Leadership in Artificial Intelligence Infrastructure, 14 January 2025
7 The White House, Executive Order on Strengthening and Promoting Innovation in the Nation’s Cybersecurity, USA, 16 January 2025
Safia Kazi, AIGP, CIPT
Is a privacy professional practices principal at ISACA®. In this role, she focuses on the development of ISACA’s privacy-related resources, including books, white papers, and review manuals. Kazi has worked at ISACA for more than a decade, previously working on the ISACA® Journal and developing the award-winning ISACA Podcast.