The AI backlash is not just online noise
June 5, 2026
by Brandon Burley and The Redemption Project
Artificial intelligence is often discussed as if the future has already arrived.
New tools.
New efficiencies.
New platforms.
New ways to write, search, automate, summarize, create, teach, diagnose, monitor, edit and decide.
But public opinion does not appear fully sold on the idea that AI and automation are making society better.
A new national registered-voter poll from Overton Insights found that only 17% of respondents viewed AI and automation as a net positive for society. Forty-six percent viewed them as somewhat or strongly net negative. Another 34% said the positives and negatives are roughly equal.
That is not a total rejection of technology.
But it is not optimism either.
It suggests voters are uneasy.
That unease deserves to be taken seriously
People may use AI and still distrust it
One of the mistakes in public conversation is assuming that if people use a technology, they must trust it.
That is not always true.
People use social media and distrust social media.
People use smartphones and worry about privacy.
People use search engines and question what they are being shown.
People use online shopping while worrying about small businesses.
The same may be true of AI.
A person may use an AI tool to write an email, summarize a document, make a video, edit audio or generate ideas while still worrying about what AI means for jobs, trust, schools, creativity, surveillance, misinformation and human judgment.
That tension is important.
Public skepticism does not always mean people want the technology banned. It may mean they want boundaries, transparency and accountability.
They may be asking a basic question:
Who controls this, and what happens if it goes wrong?
AI is not one thing
Part of the challenge is that “AI” has become a catch-all phrase.
People use it to describe chatbots, image generators, workplace automation, surveillance tools, hiring software, predictive policing systems, medical algorithms, school plagiarism detectors, customer-service bots, deepfakes, recommendation systems and more.
Those are not all the same thing.
Some tools assist human decision-making.
Some replace human labor.
Some create content.
Some analyze data.
Some sort people into categories.
Some recommend actions.
Some make decisions that affect real lives.
That is why asking whether AI is good or bad can be difficult.
A voter may think AI is useful for summarizing documents but dangerous for replacing teachers.
Another may like AI for medical research but distrust it in policing or hiring.
Another may see it as a productivity tool at work but fear what it means for their children’s future.
So when voters say AI and automation are a net negative, they may not all be reacting to the same version of AI.
That does not make the concern meaningless.
It makes the concern broader.
Work is probably at the center of the anxiety
Automation has always created public tension because it touches work.
People are not only asking whether a machine can do something faster.
They are asking whether a machine will replace them, lower their wages, monitor them, deskill their profession or make their labor less valuable.
That concern is not irrational.
Every major technological shift creates winners and losers. Some workers gain new tools. Some jobs become easier. Some industries grow.
Other workers may lose bargaining power, hours, stability or entire career paths.
When the public hears “AI and automation,” many people do not hear innovation.
They hear replacement.
That reaction may be especially strong when the benefits of new technology seem to flow upward while the disruption flows downward.
If workers believe companies will use AI to cut labor costs without sharing the gains, public skepticism will grow.
That is not only a technology problem.
It is a trust problem.
Education is another pressure point
AI also creates confusion in schools.
Teachers are trying to figure out what counts as cheating, what counts as assistance and what skills students still need to learn for themselves.
Students are trying to figure out whether AI is a shortcut, a tutor, a writing tool, a research assistant or a substitute for thinking.
Parents may not know what schools are allowing.
Administrators may be behind the technology curve.
Colleges and employers may not agree on what AI literacy should mean.
That creates a civic question bigger than classroom policy.
If AI can produce an answer instantly, what should education train students to do?
Memorize?
Analyze?
Question?
Verify?
Explain?
Create?
Detect weak reasoning?
Use tools responsibly?
The public may not have a settled answer, but the anxiety is understandable.
AI forces schools to clarify what learning is actually for.
Trust and authenticity matter
AI also raises a more human concern: What is real?
Photos can be generated.
Voices can be cloned.
Videos can be manipulated.
Essays can be produced by machines.
News summaries can be wrong.
Social media posts can be automated.
Political ads can be targeted, altered and scaled.
Scams can sound more personal.
Public comments can be flooded.
A person can hear something that sounds like a real voice and still have to ask whether it was actually said by that person.
That changes public trust.
In politics, criminal justice, education and journalism, trust matters.
If people begin to assume everything could be fake, they may stop believing true things too.
That is one of the deepest risks.
AI does not only create false information.
It can also weaken confidence in real information.
When trust collapses, civic life gets harder.
The numbers show caution across groups
The Overton poll found more negative than positive views overall, but the pattern is not simply one party rejecting AI and another embracing it.
Republicans, Democrats and independents all showed more concern than enthusiasm, though at different levels.
Independents were especially negative. Only 10% saw AI and automation as a net positive, while 62% saw them as somewhat or strongly net negative.
That matters because independents often serve as a warning light in public opinion. If they are uneasy, the issue may not stay trapped inside partisan debate.
The gender divide also matters. Women were more likely than men to view AI and automation negatively. That could reflect different workplace concerns, family concerns, safety concerns, education concerns or levels of trust in technology companies.
The poll does not explain the reasons behind every answer.
But the topline is clear: public confidence is limited.
This does not mean voters understand every detail
There is a caution here.
A poll can show public opinion without proving public expertise.
Many voters may not understand how AI systems are trained, how algorithms work, how automation differs from generative AI, what data is used, how hallucinations happen, what regulation exists or which systems are already embedded in daily life.
But that does not mean their concern should be dismissed.
Citizens do not need a computer science degree to ask whether a technology affects jobs, privacy, schools, elections, policing, medical decisions or public trust.
In a democracy, technical systems still require public legitimacy.
If people do not understand a system but are affected by it, the burden is not only on the public to learn.
The burden is also on companies, governments and institutions to explain.
Government is still catching up
AI creates a governing problem because technology often moves faster than law.
Lawmakers may not fully understand the tools they are trying to regulate.
Agencies may lack expertise.
Courts may receive cases after harm has already occurred.
Schools and employers may create policies before standards are clear.
Companies may launch tools before the public understands the tradeoffs.
That creates a gap between adoption and accountability.
Who is responsible if an AI system discriminates?
Who is liable if an automated decision causes harm?
Who owns AI-generated content?
Should political deepfakes be regulated?
Should employers have to disclose when AI is used in hiring?
Should schools require AI disclosure?
Should government agencies be allowed to use AI in criminal justice, welfare, immigration or public benefits decisions?
Those are not future questions.
They are already here.
The local angle matters too
AI can sound national or global, but its effects become local.
A school district has to decide how students may use it.
A police department may consider analytic tools.
A local business may automate customer service.
A hospital may adopt AI-assisted systems.
A city may use software to manage traffic, permits or public services.
A church, nonprofit or small newsroom may use AI to create content.
A family may deal with AI-generated scams.
The technology may be global.
The consequences are experienced locally.
That is why public unease matters for state and local government too.
If people distrust AI, they may eventually ask local officials whether it is being used in schools, courts, law enforcement, hiring, public benefits or government communications.
Those questions are fair.
The public has a right to ask when machines are being used in systems that affect people’s lives.
What this poll does not prove
The poll does not prove AI is bad.
It does not prove automation will destroy more jobs than it creates.
It does not tell us which AI tools people have used.
It does not tell us whether respondents were thinking about ChatGPT, robotics, workplace automation, deepfakes, self-driving vehicles, social media algorithms or something else.
It does not tell us what regulations voters would support.
It does not prove that public opinion will remain negative as people become more familiar with the technology.
But it does show that the public is not simply cheering from the sidelines.
People are unsure, cautious and often negative.
That is politically important.
The better way to read the result
The temptation is to say voters are anti-technology.
That would be too easy.
The better reading is that voters are not convinced the benefits of AI and automation will be shared fairly, governed responsibly or explained honestly.
That is a different point.
People may welcome useful tools.
They may still worry about who profits, who loses, who decides, who verifies and who is accountable.
That is not backward thinking.
That is civic thinking.
Technology should not be above public questions simply because it is new.
The takeaway
The AI debate is not really about machines.
It is about trust.
Do people trust companies to use AI responsibly?
Do they trust government to regulate it intelligently?
Do they trust schools to teach students how to use it without replacing learning?
Do they trust employers not to use it simply to cut people out?
Do they trust media and political institutions to protect reality from manipulation?
Right now, the Overton poll suggests many voters are not convinced.
That does not mean AI has no place in public life.
It means the public wants more than promises.
It wants proof, limits and accountability.
And that is where the civic conversation should begin.
If this kind of clear, systems-focused explanation is useful to you, The Redemption Project is where I bring together civic explainers, redemption stories, faith-centered conversations, good news features, interviews and my award-winning podcast.
You can find everything here:
Source basis: Overton Insights national poll toplines and Overton Insights crosstabs.
I am a retired detective and criminal justice / government educator based in Tennessee. I am a commentary write for Tennessee Lookout and a weekly columnist with Knox TN Today. My work examines public policy, public safety systems and civic responsibility. My reporting and commentary have also appeared in Governing, The Arizona Capitol Times, South Florida Sun Sentinel, Police1, among other state and regional outlets.
