Poll: Americans Prioritize Safe AI Development Over China Competition

A significant majority of American voters are skeptical of the notion that the U.S. should prioritize building ever more powerful artificial intelligence, unburdened by domestic regulations, in a bid to outcompete China, according to new polling exclusively shared with TIME.

These findings suggest that American voters disagree with a common narrative promoted by the tech industry, where CEOs and lobbyists have repeatedly argued that the U.S. must proceed cautiously with AI regulation to avoid giving an edge to its geopolitical rival. Moreover, the results reveal a surprising level of bipartisan consensus on AI policy, with both Republicans and Democrats advocating for government-imposed limitations on AI development to prioritize safety and national security.

According to the poll, 75% of Democrats and 75% of Republicans believe that a “careful and controlled approach” to AI—by preventing the release of tools that terrorists and foreign adversaries could exploit against the U.S.—is preferable to “pursuing AI development at the fastest possible pace to be the first nation to possess extremely powerful AI.” A majority of voters favor more stringent security protocols at AI companies and are concerned about the risk of China acquiring their most potent models, the poll indicates.

The poll was conducted in late June by the AI Policy Institute (AIPI), a U.S. nonprofit organization advocating for “a more cautious path” in AI development. The findings show that 50% of voters believe the U.S. should leverage its advantage in the AI race to prevent any country from creating a powerful AI system by enforcing “safety restrictions and rigorous testing requirements.” This is in contrast to just 23% who believe the U.S. should strive to develop powerful AI as quickly as possible to outpace China and gain a decisive advantage over Beijing.

The polling also suggests that voters may be generally skeptical of “open-source” AI, or the view that tech companies should be permitted to release the source code of their powerful AI models. Some technologists argue that open-source AI fosters innovation and diminishes the monopolistic power of major tech companies. However, others contend that it could be a recipe for danger as AI systems become more potent and unpredictable.

“Based on my interpretation of the polling data, halting AI development is not perceived as an option,” says Daniel Colson, the executive director of the AIPI. “However, granting industry complete autonomy is also seen as risky. As a result, there is a desire for an alternative approach. When we present this alternative in the polling—that third path, regulated AI development with safeguards—it is the one that people overwhelmingly favor.”

The survey also reveals that 63% of American voters believe it should be illegal to export powerful AI models to potential U.S. adversaries like China, including 73% of Republicans and 59% of Democrats. Only 14% of voters disagree.

A sample of 1,040 Americans participated in the survey, which was representative based on education levels, gender, race, and the political parties for which respondents voted in the 2020 presidential election. The margin of error for the results is 3.4% in both directions.

To date, there has been no comprehensive AI regulation in the U.S., with the White House encouraging various government agencies to regulate the technology themselves within their existing jurisdictions. However, this strategy appears to have been compromised by a recent Supreme Court ruling that restricts the ability of federal agencies to apply broad, sweeping rules established by Congress to specific or novel situations.

“Congress is so sluggish in its actions that there is considerable interest in being able to delegate authority to existing agencies or a new agency to enhance the responsiveness of government” regarding AI policy, Colson says. “This [ruling] undoubtedly makes that more difficult.”

Even if federal AI legislation appears unlikely in the near future, let alone before the 2024 election, recent polling by the AIPI and others suggests that voters are not as polarized on AI as they are on other issues facing the nation. Earlier polling by the AIPI found that 75% of Democrats and 80% of Republicans believe that U.S. AI policy should strive to prevent AI from rapidly reaching superhuman capabilities. The polls also showed that 83% of Americans believe AI could inadvertently cause a catastrophic event, and that 82% prefer slowing down AI development to address this risk, compared to just 8% who would favor its acceleration.