October 9, 2026
A new AP-NORC poll suggests that the public debate over artificial intelligence is increasingly about more than innovation. Many Americans are concerned about the speed of development, job disruption and whether policymakers can keep the technology under control.
Most Americans believe artificial intelligence is developing too quickly, according to a new poll conducted by The Associated Press-NORC Center for Public Affairs Research. The findings point to a growing challenge for technology companies and policymakers: building confidence in AI while its capabilities, commercial uses and potential effects on employment continue to expand.
The survey found that 64% of respondents thought AI was advancing too fast, while 27% said its development was proceeding at about the right pace. Around 80% emphasized the importance of keeping AI under human control and protecting workers from job displacement.
The results do not necessarily mean that Americans reject AI itself. Instead, they indicate that many want stronger safeguards and clearer answers about how the technology will affect work, privacy, children and everyday life.
The poll at a glance
Figures are from the AP-NORC poll. Percentages may not total 100% because of other responses and rounding.
Why Americans are worried about AI's pace
Artificial intelligence has moved from a specialist technology into tools used for writing, research, customer service, software development, image creation and business decision-making. Companies are also investing in systems that can carry out more complicated tasks with less direct human input.
That expansion has created a difficult public question: will the benefits arrive quickly enough, and be shared widely enough, to outweigh the risks people associate with the technology?
For some workers, the concern is that automated systems could reduce demand for certain tasks or change the skills employers need. For parents and educators, questions include how AI-generated content affects learning, how children interact with automated systems and whether personal information is being handled appropriately.
Others are concerned about inaccurate answers, misleading synthetic media and the difficulty of knowing when a decision has been made or influenced by an AI system. These concerns differ by use case; the risks of a chatbot giving a wrong answer are not identical to the risks of AI used in hiring, healthcare or financial decisions.
Jobs are a central part of the debate
AI's potential effect on employment is one of the most consequential questions facing the economy. The technology can automate portions of existing jobs, help employees complete tasks faster and create demand for new skills and services. The balance between these outcomes remains uncertain across industries.
In office-based work, generative AI can assist with drafting, summarizing documents, customer support, data analysis and coding. Employers may use those capabilities to improve productivity, but some may also reconsider how many people are needed for particular tasks.
That does not mean every job exposed to AI will disappear. Many occupations combine technical, practical and interpersonal responsibilities that cannot simply be replaced by a single software tool. The impact will depend on how companies redesign jobs, the reliability of AI systems, the cost of adoption and the need for human judgment.
The survey's emphasis on worker protection highlights a policy challenge: productivity gains do not automatically guarantee that employees will share in the benefits. Training, transitions into new roles, workplace rules and the creation of new opportunities may all influence the eventual outcome.
The central question is not only what AI can do, but how people and institutions manage its use.
Technology can increase productivity while still creating difficult choices about accountability, employment and public trust.
Americans want humans to remain in control
About eight in ten respondents emphasized keeping AI under human control. That concern becomes especially important when automated systems influence decisions that affect people's rights, finances, education or access to services.
Human oversight can mean different things depending on the application. It may involve requiring a person to review a consequential recommendation, giving people a way to challenge an automated decision, testing systems for errors or making clear when AI has been used.
Oversight is not a guarantee that every system will work correctly. A reviewer may miss an error, and an organization may rely too heavily on a machine-generated recommendation. Effective safeguards therefore also depend on transparency, independent evaluation, clear responsibility and the ability to correct problems.
Companies face a related challenge: moving quickly enough to compete while ensuring that their products are reliable and that users understand their limitations. Public confidence may become an important factor in whether people adopt AI for sensitive or high-impact tasks.
Privacy and safety add to public concerns
AI systems can process large amounts of text, images and other information. Depending on the product and its settings, users may be concerned about how their information is collected, retained, shared or used to improve a system.
AI can also make it easier to generate convincing fake images, audio, video and messages. Such content can be used to mislead people, impersonate others or make it harder to distinguish authentic material from fabricated material.
These issues are not unique to AI, but automated generation can increase the scale and speed at which some forms of content are produced. Practical protections can include limiting the information users share, checking important claims against reliable sources and using clear disclosure and reporting processes for synthetic content.
The poll's broader message is that public concerns extend beyond the workplace. People also want answers about how AI affects personal privacy, children and the reliability of information they encounter.
Do Americans trust politicians to manage AI?
The AP-NORC findings also point to doubts about political leadership. Around 40% of respondents said they trusted neither major political party to handle AI effectively, according to the poll's reported results.
The divide matters because decisions about AI are increasingly being made through a mix of legislation, agency rules, public-sector purchasing, court cases and voluntary company commitments. Policymakers must consider competing priorities, including innovation, national competitiveness, consumer protection and the economic consequences of new rules.
Too little oversight can leave people exposed to avoidable harms. Rules that are unclear, inconsistent or difficult to comply with can also create uncertainty for businesses and researchers. The policy challenge is to set meaningful protections without assuming that every AI application presents the same level of risk.
Political agreement may be difficult because AI touches several different debates at once: economic growth, education, national security, free expression, privacy and the future of work. A broad public concern about the pace of development does not necessarily translate into agreement on one specific regulatory approach.
What the poll does — and does not — tell us
Opinion polls provide a snapshot of public attitudes at a particular moment. They can show how widely a concern is shared, but they cannot establish what will happen to employment or whether a particular AI regulation will succeed.
The AP-NORC survey included 2,140 U.S. adults and reported a margin of sampling error of approximately plus or minus 2.9 percentage points. As with any survey, the wording of questions, the available response options and the timing of fieldwork matter when interpreting the findings.
The results should therefore be read as evidence of public concern, not as a forecast that AI will inevitably cause widespread job losses or that every respondent supports the same policy solution.
What happens next for AI companies and policymakers?
The findings add to pressure on technology companies to explain how their systems work, where they are appropriate to use and what protections exist when something goes wrong. Businesses adopting AI may also need to communicate more clearly with employees about changes to workflows, training and expectations.
For policymakers, the challenge is to turn broad public concerns into rules and institutions that can keep pace with a rapidly changing technology. That may involve risk-based requirements, worker training, privacy protections, testing standards and ways for people to seek explanations or challenge consequential decisions.
For workers and consumers, the immediate lesson is to evaluate AI tools based on what they can reliably do, rather than treating them as either universally beneficial or inherently harmful. Important decisions still require judgment, verification and accountability.
The survey does not settle the debate over AI. It does, however, show that public confidence cannot be assumed to grow at the same rate as technical capability. How companies and governments respond to that gap could influence the next stage of AI adoption in the United States.
Frequently asked questions
What percentage of Americans say AI is developing too quickly?
The AP-NORC poll found that 64% of surveyed U.S. adults believed AI was advancing too quickly.
How many Americans want AI kept under human control?
About 80% of respondents emphasized the importance of keeping AI under human control and protecting workers from job displacement.
Does the poll prove AI will eliminate jobs?
No. It measures public opinion, not future employment outcomes. AI may automate some tasks, change existing jobs and create new kinds of work, with effects varying by industry.
Do Americans agree on how AI should be regulated?
The findings show substantial concern, but concern does not mean people agree on specific regulations. The appropriate rules remain a matter of public and political debate.
Who conducted the survey?
The survey was conducted by The Associated Press-NORC Center for Public Affairs Research among 2,140 U.S. adults.
Sources and further reading
- The Associated Press — Most Americans think artificial intelligence is developing too fast
- AP-NORC Center for Public Affairs Research
- Newspriint — Why Washington is debating AI regulation
This article summarizes reported survey findings and general policy issues. Poll results reflect respondents' views at the time of the survey and do not predict future outcomes.
إرسال تعليق
Have something to add? Share your thoughts or let us know what you think about this story.