A Reuters/Ipsos poll found three in four Americans say AI firms are not doing enough to prevent disaster. The result signals faster regulatory action across the US and UK, while experts increasingly rank advanced AI among society’s most severe threats.
What Happened
A Reuters/Ipsos poll published September 22, 2026 found that three out of four Americans believe AI firms are not doing enough to prevent disaster. The survey captured a striking level of public unease about the direction of artificial intelligence development and the safeguards — or lack thereof — that major technology companies have in place.
The findings reflect a population that has watched AI capabilities advance rapidly while seeing limited transparency from leading labs about how they manage existential and catastrophic risks. When pollsters framed the question around disaster prevention, the three-out-of-four majority made clear that the public views corporate safety efforts as fundamentally inadequate.
The results arrived amid an already heated environment around AI governance. Lawmakers in Washington have been debating multiple legislative proposals, while regulatory bodies across the Atlantic have been crafting their own frameworks. This particular poll gives those debates a clear mandate: elected officials now hold data showing that their constituents demand action rather than promises.
What This Means For You
If you work in AI — whether as a developer, product manager, executive, or investor — this poll should change your risk calculus. Three out of four Americans expressing this level of distrust is not a temporary sentiment spike. It represents a durable shift in public perception that creates pressure translating directly into policy.
For anyone building or deploying AI systems in the United States, expect regulatory requirements to tighten faster than they have in previous technology cycles. The tech sector has historically enjoyed long windows between public concern and formal regulation. That window is closing, and this poll signals it is nearly shut. Prepare for mandatory safety evaluations, disclosure requirements, and potentially independent audit obligations before those rules fully take effect.
Business leaders who integrate AI into their operations should also pay close attention. Your vendors and partners will face new constraints, and those constraints will ripple through your supply chains. Questions about model provenance, training data handling, and deployment monitoring will become standard due-diligence checkpoints. Start asking those questions now rather than when a compliance deadline forces your hand.
Investors in AI companies should factor public sentiment risk into their models. When three out of four Americans hold this view, reputation damage can translate into revenue loss, talent attrition, and restricted market access. Companies that get ahead of the regulatory curve by demonstrating genuine safety commitment will hold a competitive advantage over those caught reacting to mandates. Consider how recent moves by major labs to release more affordable models following safety criticism signal an industry trying to respond — but whether that response satisfies the public remains an open question.
Why It Matters
This poll underscores a widening gap between the speed of AI capability development and the public’s confidence that those capabilities are being managed responsibly. That gap is not unique to the United States. The story is currently trending in the UK, where regulators have already signaled aggressive postures on AI oversight. British firms operating globally should anticipate that cross-Atlantic regulatory alignment is coming, not optional.
The scale of this concern also matters in the context of broader AI risk discourse. Recent analysis from the AI Threat Surpasses Climate Change coverage highlighted how experts are increasingly ranking advanced AI among the most severe threats facing society. The Reuters/Ipsos data reinforces that framing with concrete polling evidence: ordinary Americans are not waiting for experts to reach consensus before they demand accountability from AI companies.
For the technology industry itself, this level of public skepticism complicates one of its core narratives — that AI development is broadly beneficial and carefully managed. If three out of four Americans disagree with that assessment, the industry’s communication strategies are clearly not working. Expect messaging shifts, more safety-focused announcements, and potentially new industry self-regulatory bodies attempting to demonstrate responsiveness before governments impose their own structures. Meanwhile, enterprises exploring AI-driven risk management and stress testing should note that the public expects similar rigor from the companies building these systems — internal risk frameworks will increasingly need to match external expectations.
Key Takeaway
- Three out of four Americans say AI firms are not doing enough to prevent disaster, per a Reuters/Ipsos poll.
- Regulatory action in the US and UK will likely accelerate in response to this level of public concern.
- AI businesses should prioritize visible safety practices now, as the window between public pressure and formal regulation is narrowing rapidly.
Leave a Reply