New survey data explains How Americans use generative AI at home and at work. 57% use it for personal tasks, 21% use it on the job, and use rises with education and income. Small and large firms look similar. Productivity gains are modest, and most workers expect few new jobs from AI over the next five years. America is trying AI in everyday life. People ask chatbots to draft notes, help with homework, and search the web. Workers test AI to write documents and summarize emails. A new nationwide survey from the NORC AmeriSpeak Panel (June 2025) shows where use is strong, where it lags, and what gaps still matter. The results paint a clear picture. Personal use is common. Workplace use is early and uneven. Education and age shape adoption. Small businesses now look a lot like big firms. And many people are unsure whether AI helps their productivity or their job prospects. Below are seven findings that show what is changing and what is not.

How Americans use generative AI today

1) Personal adoption is mainstream, but it depends on education

More than half of U.S. adults have tried generative AI for themselves. In the survey, 57% report using it for at least one personal purpose. Most users turn to AI first for internet searches and browsing help. That means people see it as a fast way to find and digest information. Education is a strong driver. About 67% of adults with a bachelor’s degree or higher use AI personally. Among those with some college or an associate’s degree, 60% use it. Among high school graduates, 46% use it. This gradient shows that comfort with new tools often follows schooling. How often do people use it? Daily use still skews toward more schooling. Around one in five adults with a bachelor’s degree (20%) or some college (21%) use AI daily or more. Only 8% of high school graduates and 8% of adults without a diploma say the same.

2) Professional use is growing, but it is far from universal

Only about one in five adults (21%) use generative AI at work right now. That number climbs with education and income.

• By education: 33% of workers with a bachelor’s degree or higher use AI on the job, versus 20% with some college or an associate’s degree, 12% with a high school diploma, and 5% without a diploma.
• By age: Use peaks at 31% for ages 30–44 and stays relatively high for ages 45–59 (26%) and 18–29 (25%), then drops to 8% for ages 60+.
• By income: 9% of workers under $30,000 use AI at work, rising to 34% among those earning $100,000 or more.
• By gender: 25% of men use AI professionally vs. 17% of women.

The pattern is clear. People with more education, higher incomes, and mid-career roles are leading workplace adoption.

3) The top workplace use case is document writing and editing

When institutions support AI use, staff most often use it to draft and polish documents. This makes sense. Document work is common across many jobs and is safe to test with AI. The same gradients show up here too:

• 35% of bachelor’s degree holders use AI for documents, vs. 16% with some college, 10% with high school only, and 2% without a diploma.
• 35% of workers earning $100,000+ use AI for documents, vs. 8% of those under $30,000.
• Men report higher use than women for most workplace categories, except hiring and recruiting.

This suggests the easiest on-ramps for organizations: writing, editing, summarizing, and reformatting text. These tasks are common, low risk, and easy to measure.

4) Momentum is rising, but many workers do not know what is happening

About 22% of respondents say AI use in their workplace has increased in the last six months. Yet 61% say they are not sure, or the question does not apply to them. That “not sure” group is a warning sign. It suggests weak communication, uneven rollout, or both. Education again shows the largest gap:

• 40% of workers with a bachelor’s degree or higher report an increase in workplace AI use.
• Only 19% with some college, 9% with a high school diploma, and 5% without a diploma say the same.

Younger workers are more likely to see growth as well. 24% of adults ages 18–29 report an increase, compared to 11% of adults 60 and older. Leaders should explain where AI is live, what tools are allowed, and how they expect people to use them. That alone can reduce confusion.

5) Productivity gains are modest and hard to see

Only 19% of adults say AI has raised their daily productivity. Just 4% see a big boost. About 22% say nothing changed. More than half (53%) say they do not know or the question does not apply. Who is least sure? Older and less educated workers. Among adults 60+, 72% choose “Not applicable / Not sure.” Among adults without a high school diploma, the rate is 68%. This does not mean AI cannot raise productivity. It means many people are still early. They may not have the right use case. They may not have the skills. Or they may lack clear goals and metrics. Managers should set a baseline, pick one or two tasks, and measure changes in time and quality.

6) The job outlook is cautious

Only 11% of adults expect AI to increase job opportunities in their field over the next five years. More people believe jobs will shrink than grow. Optimism is slightly higher among lower-income workers and high school graduates (about 13%–16%) than among bachelor’s degree holders (10%) and young adults (10% for ages 18–29). This is a signal for leaders. Workers want to know how AI will change their tasks, their training, and their career paths. Clear reskilling plans, internal mobility options, and transparent policies can lower fear and raise trust.

7) Small and large firms now look surprisingly similar

The survey finds near-parity between small and large firms:

• 29% of small business respondents use generative AI in their jobs, versus 27% at larger firms.
• 59% of small business respondents and 60% of large business respondents say workplace AI use has increased in the last six months.

This suggests smaller firms have closed the early gap with big companies. Cloud tools, simple interfaces, and low-cost trials help. For many small firms, AI supports writing, marketing, customer replies, and scheduling—exactly the kind of tasks that free up hours.

Sector snapshots: health care and finance

Health care

Among health care professionals in the sample, 53% say they use AI at work. The most common application is patient communication tools (25%). Use is sharply higher among men than women (82% vs. 40%) and higher among high-income practitioners ($100,000+ in household income). Interestingly, the “some college or associate’s degree” group reports the highest use (65%), slightly above the bachelor’s-plus group (60%). This mix shows that AI is reaching both clinical and administrative roles, but adoption varies by role type and seniority.

Finance, insurance, and real estate

In a smaller sample (fewer than 50 respondents) from finance-related fields, 62% say they use AI at work. Customer service is the top use case (35%). Younger professionals report much higher use than older peers. Men also report higher use than women. One unusual finding: in this small sample, the under-$30,000 income group reports the most use. That may reflect junior roles that rely on AI for frontline support tasks.

How this study was run (in brief)

The survey comes from the NORC AmeriSpeak Omnibus, fielded in late June 2025. It includes 1,163 U.S. adults and uses a sampling design that matches the national population. Responses are weighted for selection chances and demographics. Some questions apply only to subgroups, like small business owners or health care workers. As with any self-reported survey, answers reflect what people believe about their own use. Some people may use AI without knowing it. Others may call a simple script “AI.” Keep that in mind when comparing the numbers to other studies.

How these results compare to other surveys

Other studies find similar, but not identical, patterns:

• Individual use: A 2025 YouGov study says 56% of U.S. adults use AI tools, and 28% use them weekly. Pew reports that 57% of adults interact with AI several times a week, but only 33% have used a chatbot. Another study shows that 99% of people use at least one AI-enabled product weekly, though only 36% realize it.
• Firms: Older data from 2018 shows low overall adoption but much higher adoption among very large companies. Later surveys report rising use, but the rates depend on the kinds of firms sampled. Some corporate surveys show high use among big companies, while federal data shows lower rates across all firms. This gap comes down to who gets asked.
• Small businesses: Recent reports show fast growth. One survey finds 58% of small businesses use generative AI in 2025 (up from 40% a year before). Another reports that 89% of small business owners have at least one employee using AI.
• Health care: Adoption varies by use case. Some hospitals use AI for scheduling and demand prediction, and specific diagnostic tools show sharp billing growth. Predictive models are common, but many hospitals do not yet check these models for bias or accuracy.
• Finance: Surveys show strong uptake for tasks like planning and accounting. Many finance leaders think AI will become standard within five years, but few use “agentic AI” today. Lack of trust is a top barrier.
• Government: About 42.5% of U.S. unemployment insurance agencies use AI to prevent fraud. Across public agencies, about half of employees report using AI apps several times a week.

In short, the new survey aligns with the broad story: personal use is common; professional use is growing; bigger firms had a head start; small firms are catching up; and trust, skills, and clear rules still matter.

What this means for leaders and workers

Start with simple, high-impact tasks

• Draft documents, emails, and summaries.
• Create first-pass content for marketing and customer replies.
• Clean and reformat tables and lists.
• Automate meeting notes and action items.

These use cases are safe, easy to test, and fast to measure.

Close the education and age gaps

• Offer short, hands-on training for core tasks.
• Pair less experienced users with AI “buddies.”
• Use clear prompts and shared templates to reduce friction.
• Give older workers extra time and step-by-step guides.

People learn best when they practice on real tasks they already do.

Build trust with basic governance

• Publish a simple policy: allowed tools, approved use cases, and data rules.
• Check outputs for accuracy and bias, and show how you do it.
• Protect sensitive data. Turn off training on uploaded content when possible.
• Track errors and share fixes in a common playbook.

Trust grows when guardrails are clear and issues are addressed openly.

Measure results and share wins

• Pick two metrics per use case, like time saved and error rate.
• Set a before/after baseline.
• Share simple dashboards so teams see progress.
• Reward teams that document methods others can copy.

When people see proof, adoption spreads faster.

Support small businesses with shared resources

• Use common toolkits for writing, support, and scheduling.
• Adopt vendor templates instead of building from scratch.
• Pool licenses across teams to lower costs.
• Learn from peers through local groups and online forums.

Small firms do not need heavy IT to see value. They need clear steps and low lift. The bottom line: The survey shows steady growth, clear gaps, and lots of untapped value. Personal use is ahead of workplace use. Education, income, and age shape who benefits. Small and large firms now move at similar speeds. To get more from AI, leaders should focus on simple use cases, clear rules, and real measurement. Workers should look for daily tasks that AI can speed up without hurting quality. As you plan next steps, keep one question in mind: How Americans use generative AI today shows where value already exists. The next gains will come from doing simple things well, at scale, with trust.

(Source: https://www.brookings.edu/articles/how-are-americans-using-ai-evidence-from-a-nationwide-survey/)

For more news: Click Here

FAQ