A new Reuters/Ipsos poll reveals deep public anxiety about artificial intelligence safety, with 73 percent of Americans doubting that tech companies have implemented adequate safeguards against potential harms from the technology.
The survey, released Tuesday, reflects widespread concern that AI firms have fallen short in their responsibility to prevent serious damage as the sector expands rapidly. This finding underscores a growing disconnect between the promises made by major technology companies and public confidence in their actual safety measures.
The timing of this poll matters considerably. The U.S. government, Congress, and the Biden administration have all begun scrutinizing AI development practices with increasing intensity. The White House issued an executive order in October 2023 requiring federal agencies to establish AI safety standards. Meanwhile, lawmakers on Capitol Hill have proposed multiple pieces of legislation designed to impose stricter oversight of AI systems before they reach consumers and critical infrastructure.
The survey captures sentiment at a pivotal moment for AI regulation. Tech leaders including OpenAI's Sam Altman, Google's Sundar Pichai, and others have publicly acknowledged the need for government regulation, though they often frame it in terms favorable to their own companies. This creates a credibility gap. Even when executives say they support safety measures, Americans apparently do not trust that voluntary industry efforts suffice.
The 73 percent figure represents a broad coalition of concern. It spans demographic groups and political persuasions, suggesting this issue transcends typical partisan divides. Republicans and Democrats alike worry about AI risks, though they may disagree on solutions. Some conservatives fear government overreach in AI regulation. Some progressives demand stricter controls on corporate data use and algorithmic bias.
This public skepticism directly pressures policymakers. Senator Chuck Schumer, Democrat from New York, has pushed for comprehensive AI legislation. Senator Josh Hawley, Republican from Missouri, has introduced bills focusing on different AI risks. Their different approaches reflect competing visions, but both acknowledge public demand for action.
The specific concern centers on what "serious harm" means in practice. This could include AI systems making biased hiring decisions, spreading misinformation through deepfakes, enabling cyberattacks, or causing economic disruption through job displacement. It could also involve risks to national security and critical infrastructure. Different stakeholders emphasize different dangers, but the poll shows Americans want assurances that companies are addressing all of them.
Tech companies have implemented some safeguards. OpenAI and Google have published AI safety research. Companies have created ethics boards and implemented content moderation. But these efforts remain largely self-regulatory. They lack the teeth of government enforcement or the transparency that independent audits would provide.
The survey result likely accelerates legislative momentum. Congress members citing voter concerns can justify stricter AI bills. The White House can point to this data when pressuring tech companies to accept tougher rules. Conversely, tech industry representatives will cite their own safety investments while arguing that government intervention could stifle innovation.
This public anxiety represents the real constraint on AI companies going forward. Trust erodes when 73 percent of Americans doubt your safety commitments. That gap between industry promises and public confidence will drive the next phase of AI governance in America.