84% of Americans Do NOT Expect AI to Benefit Society, Pew Research Finds

Most Americans Use AI. Most Americans Don't Trust It.

Pew Research released new data on June 17, 2026 showing a striking disconnect at the center of the AI economy: Americans are using artificial intelligence tools at rapidly growing rates, and they largely do not believe those tools will make their lives or the country better.

Only 16% of Americans think AI will have a positive impact on society over the next 20 years, according to Pew. Around 40% expect a negative impact. The remaining share lands somewhere in the neutral-to-uncertain range.

Those numbers matter because the AI industry is not some fringe experiment. It is currently driving some of the largest capital flows in American financial history. Public skepticism at this scale, running alongside record investment, is significant.

Young People Are the Most Skeptical

The conventional assumption is that younger, tech-native Americans embrace new technology fastest and most enthusiastically. For AI, that assumption is wrong.

Pew found that adults under 30 are the most negative about AI's societal impact. Only 14% of that cohort believe it will be positive. These are people who have grown up with smartphones, social media, and algorithmic feeds. They have seen what happens when transformative technology scales faster than accountability does.

When tech executives argue that skeptics simply don't understand AI well enough, keep in mind that the under-30 cohort understands it fine. They're not impressed.

Government and Companies Both Get a Vote of No Confidence

The pessimism is not limited to the technology itself. It extends to the institutions responsible for managing it.

67% of Americans do not believe the U.S. government will do anything to meaningfully regulate AI, according to Pew. 59% do not trust companies to develop it safely. Nearly two-thirds of Americans think AI development is moving too fast.

These are not fringe positions. Skepticism about federal agencies' ability to regulate fast-moving technology is a legitimate concern across the political spectrum, shared by libertarians who think bureaucracies are too slow and progressives who think regulators get captured by the industries they oversee.

The Usage Numbers Tell a Different Story

Despite the distrust, AI usage is climbing sharply.

About a quarter of Americans report using AI chatbots daily. 44% of U.S. adults now say they use ChatGPT, a figure Pew says has more than doubled since 2023. Gemini comes in second at 24%, followed by Copilot at 17%, Meta AI at 14%, Grok at 8%, Claude at 6%, and Character.ai at 3%.

Six in ten Americans routinely read AI-generated internet summaries, largely because Google has made them nearly unavoidable on search pages.

Some will read these usage numbers and conclude that the pessimism is performative. People say they're worried but keep scrolling. People use tobacco, alcohol, and social media while acknowledging those things carry real risks. Using something and endorsing its societal impact are not the same judgment.

The Strongest Counterargument

Transformative technologies have historically generated fear and backlash before delivering gains that were hard to predict in advance. Electricity, the automobile, the internet all faced serious public skepticism in early adoption phases. Pew's data captures a sentiment snapshot, not a forecast. If AI delivers meaningful productivity gains, better health diagnostics, or lower costs for goods and services over the next decade, today's 16% optimism figure could look like a historical curiosity.

This argument does not, however, answer the specific concerns Pew is capturing: that development is happening too fast for safety to keep up, that neither government nor industry can be trusted to manage it responsibly, and that the people building it are not primarily accountable to the public bearing its risks.

Gender Split

Pew also documented a consistent gender gap. Men are more likely to use AI chatbots daily — 27% versus 20% for women — and men report broader usage across multiple platforms. Women are more skeptical overall. The gap is not enormous, but it is consistent across multiple questions in the survey.

Half the Country Isn't Using It at All

About half of Americans report not using AI in their daily lives. Non-users skew older. Adults under 50 are more likely to report regular use.

This matters for policy. Any federal framework for AI regulation — or absence of one — affects the full population, including the 50% who are largely passive recipients of AI decisions embedded in hiring systems, credit scoring, medical diagnostics, content moderation, and government benefit calculations.

Pew's data was released June 17, 2026. The survey methodology and sample details are available through Pew Research Center directly. The core unresolved question the data raises: whether Congress will act on AI governance before the next major incident forces its hand, or whether the 67% of Americans who expect the government to do nothing will be proven right.