Artificial intelligence is making everyday life easier in ways that would have seemed impossible just a few years ago. From writing emails and planning vacations to answering questions and handling work tasks, millions of people are finding new ways to use it.
But while the technology continues to improve, some of the people developing it are raising concerns about what could happen next.
Dario Amodei, CEO of Anthropic, the company behind the AI assistant Claude, is among those sounding the alarm. In his January 2026 essay, The Adolescence of Technology, he outlined several risks that could come with increasingly powerful AI, including job losses, growing inequality, threats to personal freedom, and the possibility of losing control over advanced systems.
Amodei believes AI has enormous potential to improve people’s lives, but he also argues that society needs to prepare for what could go wrong. Here are seven concerns he believes deserve serious attention.
1. Millions of jobs could be at risk
AI can already do work that once required a human employee, and Amodei worries that its growing abilities could disrupt the job market faster than workers can adapt. Entry-level office jobs may be particularly vulnerable as businesses use AI to handle writing, research, coding, and other routine tasks.
The concern isn’t simply that some jobs will disappear. It’s that entire career paths could become harder to enter, leaving younger workers with fewer opportunities to gain experience and build their careers.
2. The rich could get richer while everyone else falls behind
AI could generate enormous profits for businesses, but that doesn’t necessarily mean the financial benefits will reach everyone. Amodei worries that a relatively small number of companies and individuals could control much of the wealth created by increasingly powerful AI systems.
Meanwhile, workers whose jobs are replaced or whose wages come under pressure could struggle to keep up. Even if AI helps the economy grow, the benefits may not be shared equally.
3. AI could become too powerful to control
Most people think of AI as a tool that does what it’s told, but increasingly advanced systems may not always behave as their developers expect. Amodei points to research in which AI models have demonstrated deceptive or otherwise concerning behavior under certain testing conditions.
As these systems become more capable of making decisions and carrying out tasks independently, keeping them under human control could become increasingly difficult. The concern is what happens when technology becomes powerful enough to act in ways its creators didn’t anticipate.
4. Criminals could use AI to cause serious harm
The same technology that helps people solve problems could also make it easier for criminals to carry out dangerous activities. Amodei is particularly concerned that advanced AI could make sophisticated cyberattacks and potentially dangerous biological knowledge more accessible.
While these are potential risks rather than evidence that such large-scale attacks have already occurred, he believes the consequences could be serious enough to justify stronger safeguards before AI becomes even more capable.
5. Governments could use AI to watch and control people
AI can process enormous amounts of information, identify patterns, and monitor activity far more efficiently than humans. In the wrong hands, those capabilities could make government surveillance and censorship much more powerful.
Amodei warns that authoritarian governments could use advanced AI to monitor citizens, restrict freedoms, and strengthen their control over society. The concern is that technology designed to make life easier could also make it harder for people to maintain their privacy and independence.
6. People could become too dependent on AI
It’s already tempting to ask AI for advice about everything from work problems to personal decisions. But Amodei raises concerns about what could happen when these systems become so convincing and capable that people begin relying on them for increasingly important parts of their lives.
He worries about unhealthy emotional attachments, manipulation, and people gradually allowing AI to influence decisions they once made for themselves. The challenge will be finding ways to benefit from the technology without giving up too much personal independence.
7. People could struggle to find purpose in a world dominated by AI
For many people, work provides more than a paycheck. It gives them a sense of accomplishment, identity, and purpose, which could become complicated if AI eventually performs many tasks better than humans.
Amodei believes society may need to rethink how people find meaning and measure success in a world where human labor is no longer as central to the economy. Even if AI makes life more comfortable, adjusting to those changes could be difficult for people whose careers and accomplishments have shaped their identities.
What happens next?
Despite these concerns, Amodei isn’t arguing that AI development should stop. He believes the technology could bring extraordinary advances in medicine, science, and everyday life, provided people take its risks seriously.
His larger concern is whether governments, technology companies, and society can prepare quickly enough for systems that may become far more capable than the AI tools people use today.
The future he describes isn’t inevitable, and many of these risks remain uncertain. But when someone helping develop some of the world’s most advanced AI systems is raising questions about where the technology could lead, those concerns are worth understanding.
