WHY ARE PEOPLE AFRAID OF ARTIFICIAL INTELLIGENCE NOW?

From deepfakes and job insecurity to cyberattacks and the fear of losing control, public anxiety about artificial intelligence is increasingly rooted in changes people can already see around them.

By AI TV INFO | Global Intelligence — Special Report

Artificial intelligence was once largely the subject of science fiction. Today, it writes emails, generates images and videos, assists programmers, analyzes documents and increasingly interacts with people in workplaces and everyday life.

That rapid transition has created something more complicated than either enthusiasm or rejection: a growing public unease about what happens when technology develops faster than society can adapt to it.

The concerns are wide-ranging. Workers worry about employment. Consumers worry about fraud and privacy. Journalists worry about fabricated information. Security professionals worry about misuse. Researchers debate how increasingly capable systems can remain controllable.

At the center of many of these concerns is a basic question:

How much control will humans retain as artificial intelligence becomes more capable?

A FEAR THAT IS BECOMING PERSONAL

One reason today’s AI anxiety feels different from earlier technological fears is that artificial intelligence is moving into areas traditionally associated with human judgment, creativity and identity.

Earlier waves of automation were strongly associated with factories, machinery and repetitive physical tasks. Generative AI, by contrast, can produce writing, computer code, images, music and other forms of intellectual and creative work.

That makes the technology feel much closer to people’s professional and personal lives.

Recent research illustrates the divide.

In a 2025 Pew Research Center survey, half of U.S. adults said they were more concerned than excited about the growing use of AI in daily life, while only 10% said they were more excited than concerned.

The survey also found that 56% of U.S. adults were extremely or very concerned about AI eliminating jobs, compared with 25% of surveyed AI experts.

Among employed Americans, 52% said they were worried about the future impact of AI in the workplace.

These figures do not establish that AI will eliminate particular occupations. They demonstrate something different: many people perceive significant uncertainty about what AI could mean for their economic future.

The International Labour Organization has similarly emphasized that generative AI exposure varies considerably between occupations and tasks. In many cases, AI is expected to transform or assist parts of a job rather than completely eliminate the occupation.

But uncertainty itself can be disruptive.

Workers may wonder whether their skills will remain valuable, whether wages will change or whether employers will expect them to work alongside increasingly capable machines.

WHEN SEEING IS NO LONGER BELIEVING

Perhaps the most immediate source of public anxiety is the growing difficulty of determining what is authentic.

AI can generate increasingly realistic photographs, videos, voices and written material.

The technology has legitimate applications, but the same capabilities can be used for impersonation, fraud and misinformation.

The problem is not simply that false information exists. False information existed long before artificial intelligence.

The difference is scale, speed and accessibility.

A convincing fake once required specialized equipment, technical expertise or substantial resources. Generative AI can reduce those barriers considerably.

Public concern reflects that uncertainty.

Pew Research Center reported in 2025 that 66% of U.S. adults were highly concerned about people receiving inaccurate information from AI.

In another 2025 Pew survey, 76% of Americans said it was extremely or very important to know whether an image, video or piece of text had been created by AI or by a human.

Yet 53% said they were not very or not at all confident that they could make that distinction themselves.

That creates a problem extending beyond individual scams.

If people increasingly doubt whether a photograph, recording or statement is genuine, trust itself can become harder to maintain.

For journalism, elections, public institutions and everyday communication, that uncertainty can have consequences even when the underlying content is authentic.

THE SECURITY PROBLEM

Artificial intelligence is also changing the security landscape.

AI can help defenders analyze enormous quantities of information, identify suspicious activity and respond to threats.

But similar capabilities can potentially be exploited by criminals and other malicious actors.

AI-assisted systems can help automate fraud and phishing, generate convincing impersonation attempts and assist with some forms of cyberattack.

This creates a familiar technological dilemma:

A powerful tool does not have to be inherently malicious to become dangerous when used maliciously.

The concern, therefore, is not necessarily that AI will independently decide to attack people.

Many present-day risks involve something more conventional:

humans using increasingly powerful software to operate at greater speed and scale.

That distinction is important when separating current risks from more speculative scenarios involving future autonomous AI.

THE BLACK-BOX PROBLEM

Another source of anxiety is the difficulty of explaining some AI systems’ outputs.

Large AI models learn statistical patterns from enormous quantities of data rather than following a simple collection of rules written by a programmer.

As a result, understanding precisely why a particular model produced a particular output can be difficult.

That may be relatively unimportant when an AI system is generating a recipe or summarizing an article.

It becomes more consequential when AI is involved in decisions affecting people’s lives.

Consider hiring, financial services, education, insurance or other high-stakes areas.

If an automated system produces an unexpected result, people may reasonably ask:

Why did it make that decision?

And perhaps more importantly:

Who is responsible if it was wrong?

Concerns about bias are part of this discussion.

Pew found in 2025 that 55% of both U.S. adults and surveyed AI experts were highly concerned about bias in AI systems.

The issue does not require an AI system to possess intentions or prejudices of its own.

AI systems are trained using human-generated data, and those datasets can contain existing patterns and inequalities. Systems can therefore reproduce or amplify some of those patterns.

THE FEAR OF LOSING HUMAN ABILITIES

There is another concern that is harder to measure but increasingly visible in the public debate:

What happens when people delegate more and more thinking to machines?

If AI can write the first draft, solve the problem, generate the image, summarize the book and answer the question, people may wonder whether convenience could eventually come at the cost of developing their own skills.

Pew’s 2025 research found that Americans were more likely to believe AI would make people worse rather than better at abilities such as creative thinking and forming meaningful relationships.

Among people who viewed AI’s societal risks as high, weakening human abilities and connections was a leading concern.

The fear, therefore, is not simply that machines might outperform humans.

It is that humans could gradually stop practicing some of the abilities that make human judgment, creativity and interaction valuable.

AND THEN THERE IS THE EXISTENTIAL QUESTION

Beyond today’s practical concerns lies a much more speculative category of fear:

What happens if AI eventually becomes vastly more capable than humans?

Researchers and technology leaders have debated whether sufficiently advanced AI could become difficult to control or could pursue objectives that conflict with human interests.

These scenarios concern hypothetical future systems and should not be confused with established characteristics of today’s AI.

Nevertheless, they occupy a significant place in public discussion.

Science fiction has also influenced how people imagine intelligent machines.

For decades, films, novels and television programs have depicted computers becoming autonomous, robots replacing humans and machines turning against their creators.

But concentrating exclusively on hypothetical future scenarios can obscure the risks that are already observable.

For most people, the immediate questions are more practical:

Who controls the data?

Who is responsible when an AI system makes a mistake?

Will workers share in the benefits of increased productivity?

Can people tell whether something is genuine?

And how should governments regulate increasingly powerful AI systems?

WHO IS AFRAID OF AI?

The answer is not simply “everyone.”

Different groups have different concerns.

Workers may be primarily concerned about employment, wages and changing skill requirements.

Consumers may focus on scams, privacy and the authenticity of information.

Journalists and researchers may be concerned about misinformation and the reliability of digital evidence.

Businesses may see both productivity opportunities and risks associated with cybersecurity, regulation and workforce disruption.

Security professionals focus heavily on malicious use and the changing threat landscape.

AI researchers and technology experts debate both present-day risks and longer-term questions about the control of increasingly capable systems.

And the broader public may experience several of these concerns simultaneously.

That makes AI anxiety less like a single movement and more like a collection of different fears surrounding one rapidly developing technology.

THE QUESTION OF CONTROL

The common thread running through many of these concerns is ultimately control.

Pew Research Center reported in 2025 that 55% of U.S. adults wanted more control over how AI was used in their lives.

The same research found that majorities of both the public and surveyed AI experts were more concerned that government regulation would be insufficient than that regulation would go too far.

At the same time, public concern does not necessarily mean rejection of artificial intelligence.

People continue to use AI and recognize potential benefits in areas ranging from research and education to accessibility, programming and everyday productivity.

Pew also found that nearly three-quarters of Americans surveyed would be willing to let AI assist them at least somewhat with everyday activities.

That creates a striking paradox.

People can simultaneously find AI useful and worry about its consequences.

THE BOTTOM LINE

The public debate over artificial intelligence is not simply a debate about whether machines will become dangerous.

It is increasingly a debate about speed, power and responsibility.

AI is developing rapidly. Its applications are spreading into workplaces, media, education and everyday communication. Meanwhile, governments, institutions and individuals are trying to determine how to adapt.

The fears are therefore not all about a hypothetical future.

Some are already visible:

job uncertainty, deepfakes, scams, privacy concerns, bias, cyber threats and uncertainty over who is responsible when AI goes wrong.

Other fears remain speculative, particularly those involving hypothetical systems far more capable and autonomous than today’s technology.

But both categories point toward the same underlying question: