With a presidential election just days away, business leaders are considering how the result will impact AI plans and existing practices.
Enterprises are trying to keep up with a patchwork of regulations, with more than half of technology leaders saying that complying with new legislation presents a significant challenge, according to PwC.
The U.S. holds a unique position in shaping global AI standards as home to many of the world’s leading AI companies and developers. But analysts say partisan disagreement and the current patchwork of state rules are hurting the nation’s ability to influence AI oversight globally.
“Without a unified federal approach, the U.S. will find it challenging to lead globally in AI regulation,” Bobby Malhotra, partner at Winston & Strawn and part of the firm’s AI strategy group, said in an email.
The next president will take office at a time when businesses are looking for regulatory clarity that mitigates the risks of the technology while enabling innovation.
Vice President Kamala Harris and former President Donald Trump have said they will support policies that encourage homegrown innovation if elected, but they have carved different paths to reach the goal.
The Harris-Walz platform said its pursuit of AI governance will build upon President Joe Biden’s AI executive order, issued in October 2023. The Trump-Vance policy proposal calls for a repeal of the provisions. Neither campaign responded to a request for comment.
“Congress is not likely, under any administration, to actually congeal around a unified stance on AI,” Russell Newman, associate professor at Emerson College and faculty associate at Harvard University, told CIO Dive. “But if we’re not talking about regulation, and we’re just talking about ‘winning the AI race,’ you could do it with whoever gets elected.”
A lighter regulatory touch could drive more AI development, but experts also worry about the lack of obligatory guardrails to prevent misuse.
“You have to be very thoughtful around how the legislation comes together so it doesn’t stifle innovation and creation,” said Ari Lightman, distinguished service professor at Carnegie Mellon University. “But at the same time, you have to have checks and balances, otherwise there will be runaway catastrophic events that occur.”
Around the world, legislators have taken aim at AI. The European Union’s AI Act is expected to have far-reaching implications with its guardrails on developing and deploying AI for organizations operating in its jurisdiction. Other countries, including China, Canada and Brazil, have advanced or are in the process of developing comprehensive bills to regulate high-risk systems as well.
The U.S. is expected to propose legislation, pass laws at the state level, provide industry standards and influence AI policy, but not as cohesively as the EU, Jennifer Everett, partner in Alston & Bird’s technology and privacy group, said in an email.
AI in the backseat, but unlikely for long
Despite its far-reaching implications and wide array of use cases, AI has not gained a leading role in campaign speeches and candidates’ policy proposals.
“People have a mushy understanding of what AI is and how AI works,” Vincent Raynauld, associate professor at Emerson College, told CIO Dive. He expects AI to gain more importance within political platforms in the next few years.
Still, AI’s myriad of use cases means sector-level initiatives and policies could impact governance of the technology, fueling the need to introduce comprehensive guardrails.
“Explicit AI policy per se certainly isn’t in the forefront, but it’s there in the background,” Newman said. “It’s there in how we discuss the future of manufacturing and the future of working life overall.”
The U.S. Department of Labor published a comprehensive guide of best practices and principles for AI developers and deployers to better worker well-being earlier this month, after being tasked by the Biden administration’s executive order. The Harris-Walz campaign plans to build upon the commitments laid out in the AI executive order but also promised to invest in “cutting-edge” technologies to support American manufacturing.
The Trump-Vance campaign proposal highlights the use of “advanced technology” to support border and military initiatives.
“There are many other high-profile issues on the table, but AI is embedded in a lot of the conversations,” Newman said.
Historical precedents also suggest that comprehensive legislation on AI could come further down the line. The General Data Protection Regulation went into effect in the EU in 2018. While U.S. states have passed privacy laws following GDPR, the country still lacks a comprehensive federal privacy law.
Significant obstacles to leading globally on AI oversight remain despite the ambitions expressed by Harris and Trump, according to Malhotra.
The deep-seated partisan disagreements within the U.S. government create “substantial challenges in achieving consensus on comprehensive federal legislation, thereby potentially thwarting the possibility for the U.S. to lead in this space,” Malhotra said.
Technology leaders still have to keep up with the fast-paced regulatory environment abroad and stateside, no matter who wins next week.
“Regardless of the presidential election, businesses should continue to be mindful of ongoing legislative developments and regulatory enforcement at the state level,” Everett said.
