Bipartisan Support Grows for Government Regulation of AI Amid Rising Concerns

A recent study conducted by the Program for Public Consultation (PPC) at the University of Maryland underscores a growing bipartisan consensus among Americans urging the government to take a more active role in regulating artificial intelligence (AI). Despite the White House's recent AI Action Plan advocating against intervention due to potential stifling of innovation, survey results show substantial public backing for specific regulatory measures. This insightful data was gathered from a nationwide survey of 1,202 adults conducted between July 30 and August 7, 2025.

The Public's Concerns About AI

The backdrop for this study highlights significant apprehensions regarding the implications of uncontrolled AI development. Steven Kull, the director of PPC, emphasized that while the public is cautious about excessive government regulation, their anxieties regarding unrestrained AI growth are even more pronounced. The survey reveals that Americans from both political spectrums recognize the necessity of implementing regulatory measures to safeguard against potential risks, particularly in decision-making contexts.

Emerging Regulatory Proposals

The findings reveal an overwhelming majority of respondents favoring five key proposals for government regulation:

1. Pre-deployment Testing: 79% support the idea that AI programs be evaluated and approved by a government body to ensure compliance with laws and standards, reflecting high levels of concern about biased or harmful outcomes in various fields, including healthcare and criminal justice. This support spans across party lines with 84% of Republicans and 81% of Democrats in favor.

2. Audits of Existing AI Programs: Similarly, 78% support allowing government audits of currently deployed AI systems, with bipartisan agreement that accountability measures should be in place to rectify any identified issues, such as discrimination and security flaws.

3. Transparency in AI Training: The demand for companies to disclose training data of AI systems has gained traction, with 74% endorsing this requirement to facilitate effective audits and ensure data integrity.

4. Regulating Deepfakes: There is also significant support (around 80%) for prohibiting the use of AI-generated deepfakes in political campaigns, with widespread agreement that misinformation undermines electoral integrity. Respondents expressed alarm about the potential repercussions of deceptive media, reinforcing the necessity for labeling and transparency.

5. International Treaties on Autonomous Weapons: Importantly, the study highlighted bipartisan support (74%) for an international treaty to ban lethal autonomous weapons, despite contrasting governmental positions. Respondents acknowledged the dangers of AI applications in warfare and supported collaborative efforts to create a safer global environment.

Public Sentiment on AI Regulation

This data calls attention to a societal desire for proactive measures when considering AI's rapid advancement. Although arguments against rigorous regulatory frameworks emphasize innovation concerns, the results illustrate that a larger segment of the population prioritizes safety and accountability.

The survey juxtaposes apprehensions about potential overreach against the need for systematic checks to maintain ethical standards in AI deployment. Public sentiment appears to resonate with the notion that precaution should guide AI development rather than reactive measures post-failure.

Conclusion

This study from the University of Maryland signals a compelling mandate for policymakers to take heed of public concerns and facilitate discussions regarding the future of AI regulation. As society navigates the complexities of technological advancement, fostering cooperation and establishing regulatory frameworks will be pivotal to ensure beneficial outcomes for all. The findings advocate for integrating public opinion in shaping legislation that governs AI technologies, reflecting the genuine stakes involved in its deployment.