What will the US election mean for AI regulation?

The upcoming US presidential election promises to be one of the most hard-fought and contentious in recent history.

And with incumbent President Joe Biden stepping down just months before Americans go to the polls, this year’s election has already entered unchartered waters.

Depending on where you get your news, you might consider the US election pivotal for the future direction of the economy, trade and industrial policy, abortion rights – or even for democracy itself.

But perhaps understandably amid the election furore, few commentators have cast a serious glance towards what the US election might mean for the direction of AI policy.

And while you might believe that AI regulation pales in importance to the other issues at stake, the industry’s potential to revolutionise economic productivity and transform how we interact with technology means getting AI policy right is one of the major asks the next administration must get right.

The Tech announcer spoke to Rotem Farkash, cybersecurity expert and entrepreneur with years of experience navigating AI regulation, who talked us through the current state of US AI policy, and what a Harris or Trump victory might mean for the industry.

State of play

Farkash immediately highlights that, as of the time of writing, the US has ‘no comprehensive federal legislation’ to specifically prohibit or restrict the use of AI.

Instead, President Joe Biden signed a sweeping executive order instructing the government to develop standards for labelling online AI content, and setting out guidelines for companies to ‘red team’ the AI products, where assessors emulate rogue actors to uncover any possible exploitations of the technology.

Acting under the 1950 Defense Production Act, Biden’s order further required companies to disclose safety test results with the government when developing AI models that may pose a threat to national security.

There are limits to the President’s executive power, however, and as Farkash explains, the deadlock between the President and a Republican-controlled Congress threatens to handcuff the US response to AI at a critical time in the industry’s development.

And as well intentioned as Biden’s order was, the use of emergency powers to compel private companies to row back on innovation by designating AI a national security threat has certainly ruffled feathers, with industry figures describing the order as ‘a clear violation of executive authority’.

This is in sharp contrast with the US’ two key rivals in the sphere, China and the EU, who have already imposed controls on AI over fears of the consequences of unregulated AI research.

China have been quick to ensure that AI development takes place under government oversight, with the Financial Times recently reporting that the Chinese regulator, the Cyberspace Administration of China, has forced AI start-ups to undergo a mandatory government review to ensure large language models embody ‘socialist values’.

The US is more likely to look towards the EU for inspiration, who passed a comprehensive Artificial Intelligence Act in May of this year which imposes strict transparency restrictions on high-risk AI systems while heavily restricting the use of AI in biometric surveillance.

But while commending the bloc’s concern for the undoubted risks unregulated AI development poses, Farkash points to the ongoing regulatory uncertainty which has led Meta and Apple to suspend the launch of new AI products in Europe, highlighting the risk of future disputes.

‘The EU will likely come under pressure from large corporates, as well as customers, to relax AI regulation, who will compare tech products on offer with those available to their American counterparts. The EU will have to navigate difficult compromises to achieve a balance acceptable to businesses and consumers,’ Farkash told the Daily Caller.

What’s at stake during the election?

Since both major parties have all but confirmed their Presidential candidates for November, the Daily Caller asked Rotem Farkash to speculate on what US AI policy may look like under a Harris or Trump presidency.

Beginning with the Democrat candidate, Farkash points to the leading role Kamala Harris has played in driving the current administration’s policy on AI,  convening a meeting with the CEOs of Google, Microsoft, OpenAI and Anthropic in May 2023 to discuss approaches to AI regulation where they discussed responsible approaches to AI innovation.

Although the results of this meeting were unclear, it points to a desire to engage constructively with industry leaders to lay out standards of corporate responsibility in AI.

But as a former attorney general of California, Harris had a track record in advocating for stricter regulation on big tech, pushing for laws against cyberbullying, and calling for big tech firms to take down intimate images on social media posted without their subject’s consent.

‘It is likely a Harris administration will take a more hands-on approach to AI regulation, although she made steps to reassure big tech companies that she is on the side of innovation,’ Farkash said.

In contrast, senior Republican figures have committed to liberalising restrictions on AI, and the Party has already pledged to reverse Biden’s executive order of 2023.

Trump’s running mate JD Vance, meanwhile, is a known sceptic of AI regulation, having accused big tech firms of promoting new regulations only to entrench their positions in the market.

‘It’s still early days, and the Republicans have yet to make firm pronouncements on what their AI policy will look like. But pursuing the laissez-faire policies on AI that are widely expected will put the US at odds with the EU’s more interventionist approach. Differences in regulatory regimes between the world’s two largest markets will pose significant challenges to AI entrepreneurs,’ Farkash explained.

In his concluding remarks, Farkash stressed that AI entrepreneurs should pay particular attention to the first few months of the new administration: ‘AI regulation is very much unchartered territory, and the differences between an interventionist and laissez-faire approach to AI policy are significant. Entrepreneurs will have to react quickly to the approach the next US administration chooses to take.’