As AI companies race to build more powerful models and enormous computing infrastructure, a different constraint is beginning to emerge: a growing share of the public is becoming more worried about artificial intelligence even as the technology becomes more capable.
WHAT’S HAPPENING
Artificial intelligence continues to spread rapidly through workplaces, search, education, healthcare and consumer products. But public enthusiasm is not necessarily keeping pace.
New Pew Research Center data shows that 55% of Americans ages 18–29 now say they are more concerned than excited about the increasing use of AI in daily life, up from 31% in 2021. Only 11% of that age group now say they are more excited than concerned.
Concern about employment is even broader.
71% of U.S. adults believe AI will lead to fewer jobs over the next 20 years, compared with 64% in 2024. Among adults under 30, the number has climbed to 73%.
The skepticism extends beyond jobs.
Communities have pushed back against data-center development, creators continue debating compensation for material used to train AI models, and concerns remain about AI’s effects on education, creativity and human relationships.
The backlash has even entered popular culture. Liquid Death and Garage Beer recently teamed with former NFL star Jason Kelce on a deliberately absurd advertising campaign suggesting people send urine to AI data centers to help cool them — turning concerns about data-center water consumption into mainstream comedy.
Meanwhile, some leaders inside the AI industry are beginning to acknowledge the larger problem.
Anthropic CEO Dario Amodei recently described the backlash surrounding AI as fundamentally a “crisis of trust.”
WHY IT MATTERS
For years, the biggest constraints on AI expansion appeared to be technical and physical:
chips, compute, electricity, data, water and capital.
But there may be another one.
Public acceptance.
An AI system can become extraordinarily capable without automatically convincing workers, parents, creators, communities or consumers that its expansion will benefit them.
And simply changing the message may not be enough.
A recent Searchlight Institute experiment tested different ways AI leaders talk about the technology. Neither highly alarming rhetoric nor a more optimistic description meaningfully improved attitudes toward AI.
That suggests the industry’s challenge may extend beyond branding.
People may increasingly judge AI based on what they actually experience — whether it improves their work, threatens their opportunities, affects their communities or creates benefits they can clearly see.
WHO BENEFITS
Companies that can establish trust alongside capability could gain an important competitive advantage.
That may favor AI businesses that demonstrate clear usefulness, protect user data, explain how their systems work, give people meaningful control and show tangible benefits rather than relying exclusively on promises about what AI could eventually accomplish.
Governments, educators and organizations that help people understand how AI is being used may also become increasingly important as adoption expands.
And companies that successfully use AI to augment workers rather than simply presenting it as a replacement for them may have an easier path to acceptance.
WHO LOSES
AI companies that assume better technology automatically produces greater acceptance could face resistance even when their products work.
Data-center developers may encounter community opposition.
Employers could face worker resistance if AI deployments are associated primarily with job elimination.
And companies that fail to establish trust around privacy, transparency, intellectual property or safety may discover that technical superiority alone does not guarantee widespread adoption.
The biggest risk may be a widening gap between what AI can do and what society is comfortable allowing it to do.
WHAT HAPPENS NEXT
Watch whether public attitudes begin improving as AI becomes more useful in everyday life — or whether concern continues rising as the technology becomes more powerful.
Also watch the behavior of the AI companies themselves.
The next phase of competition may not be fought only over who has the best model.
It may increasingly involve who can convince people that deploying that model creates more value than disruption.
Silicon Valley has become extremely good at scaling technology.
Its next challenge may be scaling trust.
