Skip to content
AI Survey 2026 Photo

Survey

AI Is Getting Real, But the Real Work Is Still Ahead

Explore the Findings

Introduction

New results from a Jobs for the Future (JFF) survey shed light on how workers and learners perceive, use, and experience the impacts of AI at work, in education, and in career navigation.

Our findings include:

  • A “messy middle” moment of AI advancement and adoption that reflects AI’s uneven capabilities.
  • A lack of employer training and guidance on AI tools and adoption.
  • Sharp divisions over AI’s impact and potential for good or harm.

For those who use AI, the benefits are many and moving toward augmentative impacts, but a large portion of respondents still do not use the technology. AI’s presence in work and learning is well underway, but the real work—integrating it into how people do their jobs and live their lives—is just beginning.

Fragmented attitudes and adoption

The sample of more than 3,000 Americans included an over-sample of people without a four-year college degree, people of color, even those with a four-year degree, women, even those with a four-year degree, and people with a record of arrest, conviction, or incarceration; this ensures that the perspectives of members of these populations are included to a greater degree than they are in the mainstream discourse around AI.

Respondents revealed that workers’ concerns about and adoption of AI are highly fragmented. To gain information and skills in AI, they are turning to nontraditional sources (including YouTube tutorials and personal experimentation) rather than programs offered by education and training providers, and they say they find it difficult to locate high-quality instruction. Critically, workers say they aren’t receiving sufficient training, preparation, and consultation from their employers regarding AI.

On topics related to navigating AI-related career shifts and training needs, workers’ responses reflect the lack of a clear, large-scale impact of AI on the labor market and the ambiguity around what the future holds. Respondents are sharply divided on whether AI is impacting their job search and when and how urgently they may need to take action. They overwhelmingly recognize that the skills required for jobs are shifting, but many say they don’t see the need to gain new skills. People of color and early-career workers in particular say they feel the impact of AI on their jobs and report that they’re becoming more proactive about acquiring new skills and making career shifts. Women, who are already concentrated in jobs at risk of automation, also express less optimism about AI and say they’re less interested in increasing their use and knowledge of AI—outlooks that may put them at risk of being left behind.

For JFF, these findings underscore both the promise and the risks related to the potential for economic advancement in an increasingly AI-powered world. Employers, education and training providers, and policymakers need to act now to ensure that we don’t leave anyone behind:

  • Employers should invest in training and structured support for all workers to ensure that they understand AI and are able to use it in the workplace; they should also consult with workers directly about AI use on the job.
  • Education and training institutions shouldn’t try to compete with the abundant AI instructional content that’s available online; instead, they should differentiate themselves by promoting the fact that they offer high-quality opportunities to build AI skills in context through work-based learning.
  • Policymakers and intermediaries should ensure consistency in the implementation of AI and access to AI and AI-related learning experiences so that all people can reap the benefits of this technology.
  • Organizations that provide career navigation services need to prepare to better support jobseekers in making informed decisions about AI training and career opportunities.

Below are additional highlights from the survey, which was conducted in November 2025.

NotebookLM runs on a generative AI model and may make mistakes.

Overall Trends

0%

of respondents reported that AI tools had at least “some impact” on their jobs, up from 57% in 2024.

0%

reported that “AI has not impacted my responsibilities at work,” signaling uneven diffusion.

0%

said they believe AI is currently having a great deal of impact on education, an increase from 33% in 2024.

0%

said they felt optimistic about AI’s future impact on education, a sharp decrease from 51% in 2024.

AI is having a bigger impact, and concerns are on the rise.

  • In our 2025 survey, respondents who said AI is doing more harm across society as a whole outnumbered those who said it was doing more good (44% said more harm; 38% said more good). This represents a notable reversal from our 2024 survey.
  • Around half (49%) of women said AI is doing more harm than good, while only 33% said AI is a net-positive for finding jobs, building wealth, and securing their quality of life.
  • Women expressed less optimism than men about AI’s impact on jobs (31% reported that they’re “somewhat” or “very” optimistic, versus 47% of men) and education (36% versus 52%). Women were also less likely than men to express interest in increasing their familiarity and use of AI tools (31% vs. 44%).
  • Top-cited concerns about AI harms include overreliance on AI and/or loss of human control (46%), security risks (47%), job displacement and/or changes to wages and work (45%), and privacy concerns (45%). Learners in particular most often cited concerns about privacy (51%), job displacement and/or changes to wages and work (39%), security risks (41%), and overreliance on AI and/or loss of human control (42%).

Workers and learners are turning to nontraditional sources for AI information and training.

  • Respondents most commonly said they seek information about AI from social media (31%), news articles (27%), and friends and family (21%). By comparison, 9% said they seek information from their employers, 8% said they seek formal education, and 8% said they turn to local workforce boards.
  • Respondents looking to increase their familiarity with and use of AI tools most commonly said they plan to experiment on their own (62%) or watch YouTube tutorials or informal courses (53%). Taking a free course from a local organization (25%) and enrolling in a paid course through a college or university (16%) were much less common responses.
  • Over one-third of respondents (37%) said they want to increase their familiarity with and use of AI tools, a decline from 48% in 2024.
  • Respondents cited limited time (37%), the high cost of courses and tools (33%), and difficulty finding reliable sources of information (30%) as the top barriers for increasing familiarity with and use of AI tools.

The impact of AI on career trajectories is mixed at this early stage. 

  • Only 27% of respondents said they have turned to AI for career advice, and 14% report that this advice has significantly influenced their career decisions.
  • 46% of respondents said they believe AI has not impacted their ability to get a job, 26% said they believe AI has made it more difficult, and 28% said they believe AI has made it easier.
  • A total of only 23% of respondents said that they are either actively changing career paths (11%) or are planning to change career paths in the near future (12%) as a result of AI—a slight increase from 2024 (19%, 7%, and 12%, respectively).

Do you believe AI is having more harm than good, or more good than harm on people’s ability to find jobs, build wealth, and secure their quality of life, in society as a whole?

Bar chart comparing perceptions of whether AI does more harm than good in society in 2024 and 2025. In 2024, 45% of respondents said AI does more good than harm and 41% said it does more harm than good. In 2025, opinions shifted, with 44% saying AI does more harm than good and 38% saying it does more good than harm.

How do you think AI has impacted your ability to get a job?

Bubble chart showing how AI has affected people’s ability to get a job. Forty-six percent of respondents say AI has not impacted their ability to get a job, 28% say AI has made it easier to get a job, and 26% say AI has made it more difficult to get a job.
Bubble chart showing how AI has affected people’s ability to get a job. Forty-six percent of respondents say AI has not impacted their ability to get a job, 28% say AI has made it easier to get a job, and 26% say AI has made it more difficult to get a job.

AI at Work

0%

of workers reported having access to employer-provided paid AI tools.

0%

of workers said that their employers currently offer training on AI skills (up from 31% in 2024).

0%

of workers reported that they have the training and resources they need to use AI in their jobs, down from 45% in 2024.

Workers still don’t seem to be getting enough support from their employers around AI.

  • Similar to 2024, respondents to the 2025 survey were more likely to say that they use AI at work on their own initiative. However, we did see an increase in respondents saying they adopted AI with their employer’s direction (from 11% to 18%).
  • People without four-year degrees were less likely to say they have received AI training at work (31% vs. 37% total); and among respondents without four-year degrees, women were less likely than men to say they have received AI training at work (34% versus 42%).
  • 56% of workers said they have not been consulted by their employers on how AI tools are used in their work. 24% said they have served actively on an AI committee or similar group, and 19% said they have been asked for feedback.
  • People of color were most likely to say they had been consulted by their employers about the use of AI in the workplace (51%). In comparison, women (39%), people without four-year degrees (36%), and people with records (40%) were less likely to say they had been consulted.

Has your employer consulted you in some way about how AI tools are used in your work?

Horizontal bar chart showing whether workers have been consulted by their employer about how AI tools are used at work. Fifty-six percent say their employer has never consulted them, 24% say they are actively involved in discussions, and 19% say their employer has asked for their feedback.

How prepared do you feel to use AI effectively in your job?

Stacked bar chart comparing workers’ preparedness to use AI effectively in their jobs in 2024 and 2025. In 2025, 17% say they are not prepared at all, 7% not very prepared, 40% neutral, 19% somewhat prepared, and 17% very prepared. In 2024, 10% said they were not prepared at all, 11% not very prepared, 35% neutral, 25% somewhat prepared, and 20% very prepared.
Stacked bar chart comparing workers’ preparedness to use AI effectively in their jobs in 2024 and 2025. In 2025, 17% say they are not prepared at all, 7% not very prepared, 40% neutral, 19% somewhat prepared, and 17% very prepared. In 2024, 10% said they were not prepared at all, 11% not very prepared, 35% neutral, 25% somewhat prepared, and 20% very prepared.

How are you currently using AI tools to complete your work and job responsibilities?

Horizontal bar chart comparing how workers used AI tools for different work tasks in 2024 and 2025. In 2025, the most common uses were research purposes (41%), analyzing data (39%), learning and training (38%), innovation (37%), and generating content (32%). Other uses include automating repetitive tasks (34%), project management (33%), communication support (31%), supporting customer interactions (29%), enhancing decision-making (28%), and coding or software development (20%). Most categories show increased use compared with 2024.

The frequency of AI use has plateaued since 2024, but AI users are turning to more strategic use cases and starting to see augmentative benefits that enhance human capabilities. 

  • 38% of workers reported using AI at work, compared to 35% in 2024. 36% of workers reported that they rarely or never use AI for any use—personal or work.
  • Reported uses of AI for research (41%), learning and training (38%), and innovation (37%) at work all increased by at least 5 percentage points from 2024, while reported uses of AI to generate content declined from 37% to 31%.
  • In 2025, workers were more likely to report benefits of using AI related to augmentation, including creating new types of tasks or responsibilities that did not exist before (13% vs. 7%), increasing the complexity of tasks (12% vs. 8%), and making work more collaborative (14% vs. 10%).

Most workers are recognizing significant changes to skill needs and seeing the need to upskill as a result of AI. 

  • Workers said they see an increase in the importance of technical skills (38%), problem-solving abilities (40%), adaptability (38%), and strategic thinking and decision-making (37%) in their jobs. Only 7% said they do not think AI is significantly changing the importance of any skills (a drastic decline from 42% in 2024).
  • In recognition of these skills shifts, workers said they’re feeling the need to acquire new skills due to AI's impact on their work (47%, with 29% of respondents saying they need to acquire new skills within the next year). Still, more than half of workers (52%) said they don’t feel the need to develop new skills.
  • Similar to 2024, people of color responding to the 2025 survey were much more likely to report using AI and to say they feel the urgency of upskilling and career changes as a result of AI. 38% of people of color said they’re actively changing or planning to change their career pathways in the near future as a result of AI (compared with 23% overall), and 44% said they feel the need to gain new skills in the next year (compared with 29% overall).

How do you feel AI is impacting the skills you need to be successful in your job?

Horizontal bar chart comparing how workers used AI tools for different work tasks in 2024 and 2025. In 2025, the most common uses were research purposes (41%), analyzing data (39%), learning and training (38%), innovation (37%), and generating content (32%). Other uses include automating repetitive tasks (34%), project management (33%), communication support (31%), supporting customer interactions (29%), enhancing decision-making (28%), and coding or software development (20%). Most categories show increased use compared with 2024.

Workers in early stages of their careers were more likely than workers with more experience to say they have a good understanding of AI. Their responses also indicated that they feel the impact of AI on their jobs more acutely than older workers, and that they are more likely to see a need to upskill or change careers because of AI.

  • Early-career workers (those with 0-3 years of experience) were more likely than those with more experience to report having a good understanding of AI (69% vs. 45% of all others); they were also more likely to say they feel prepared to use AI at work (51% vs. 33%).
  • Early-career workers were more likely than older workers to express concerns about overreliance on AI (49% vs. 46%).
  • Early-career workers were more likely than workers with more experience to report that AI has had some or a great deal of impact on their jobs (74%  vs. 64%).
  • Early-career workers were more likely than older workers to say it is more challenging to get a job because of AI (27% vs. 25%), and workers with less than one year of experience were most likely to report difficulty (39%).
  • Workers with less than three years of experience were much more likely than older workers to say they had changed or are considering changing their career plans in the near future as a result of AI (40% vs. 19%), and they were more likely to say they feel the need to gain new skills within the next year as a result of AI (39% versus 27%).

Has the impact of AI influenced the career paths or work you are considering for the future?

Stacked horizontal bar chart showing whether AI has influenced workers’ future career plans, comparing all workers, early-career workers (0–3 years), and later-career workers (3 or more years). Overall, 55% say they do not believe they need to change career paths because of AI, 14% are unsure, 12% are considering changing careers in the near future, 11% have already changed plans, and 8% are considering changes in the longer term. Early-career workers are more likely to consider or pursue career changes due to AI than later-career workers.
Stacked horizontal bar chart showing whether AI has influenced workers’ future career plans, comparing all workers, early-career workers (0–3 years), and later-career workers (3 or more years). Overall, 55% say they do not believe they need to change career paths because of AI, 14% are unsure, 12% are considering changing careers in the near future, 11% have already changed plans, and 8% are considering changes in the longer term. Early-career workers are more likely to consider or pursue career changes due to AI than later-career workers.

AI and Education

Use of AI and support for learners in the classroom is on the rise. 

  • 69% of learners said they had received AI training from their educational or training institution, up from 47% in 2024. In the 2025 survey, 35% of those learners reported that training was highly effective, up significantly from 15% in 2024. Meanwhile, 14% of learners reported receiving no training (down from 38% in 2024).
  • 70% of learners reported using AI for their education at least weekly (up from 59% in 2024), while 22% said they rarely or never use AI tools. 69% of learners reported AI tools being incorporated into their lessons or training (up from 57% in 2024).
  • When asked why they are not frequently using AI tools in their education or training, learners cite perceived lack of value (54%), lack of trust (53%), and a preference for traditional learning models (53%).
  • Similar to 2024, learners were more likely to say they had used AI for education on their own initiative (18%). However, there was an increase in respondents who said they had adopted AI at their education or training institution’s direction (from 8% to 11%).
  • 37% of learners reported having access to school-provided paid AI tools (compared with 17% of all respondents)—a finding that's likely driven in part by programs that provide students with free access to AI from the frontier model companies.
  • The percentage of learners who reported that their education or training institution fully permits and encourages the use of AI tools increased significantly—from 10% in 2024 to 31% in 2025. Meanwhile, 11% of learners reported that their education or training institution bans AI use completely, and 13% said they don't know what the policy is.

Have you received any training on how to use AI tools in your education from your education or training provider?

Stacked bar chart showing whether learners received training on how to use AI tools from their education or training provider and whether the training was effective. In 2024, 35% report receiving training. In 2025, 56% report receiving training, including 35% who say the training was highly effective and 21% somewhat effective. Training access varies across groups: 53% of people without a four-year degree, 56% of people of color, 51% of women, and 79% of people with records report receiving training.

Learners are most commonly using AI to enhance learning and are finding benefits from efficiency gains.  

  • Learners reported using AI for a variety of purposes, including completing assignments more efficiently (44%), enhancing learning and understanding (38%), exploring additional learning resources (36%), and gaining access to AI-driven tutoring or assistance (35%).
  • 25% of people of color reported being active daily users of AI tools in education, particularly for research (47%) and presentations (40%).
  • Asked to name the perceived benefits of using AI in education, learners’ most common responses were helping to understand complex topics more easily (30%) and improving efficiency in completing assignments (29%).
  • 11% of learners said they believe AI has not significantly supported their education, pointing to a need for additional support for some learners.
  • 45% of learners say they want to increase their familiarity and use of AI tools, but 41% do not. Learners cite limited time (36%), high cost of courses and tools (40%), and a prohibition on using AI in their education or training programs (31%) as top barriers.
  • Despite increased institutional training, learners are still more likely to seek AI information from informal channels, including social media (48%), news articles (38%), and friends and family (30%), compared to their school or training program (23%), workforce boards or career centers (23%), or conferences (8%).
  • When looking to increase familiarity with AI tools, learners report taking a largely self-directed approach, including experimenting on their own (46%), using YouTube or informal courses (44%), enrolling in a paid college or university course (43%), or taking a free local course (31%).

How often do you currently use AI tools in your education or training?

Stacked bar chart showing how often learners use AI tools in their education or training. In 2024, 59% report using AI weekly or daily, including 25% daily and 34% weekly. In 2025, this increases to 70%, including 22% daily and 48% weekly. Usage is also high among specific groups, including 72% of people without a four-year degree, 72% of people of color, 66% of women, and 78% of people with records reporting weekly or daily use.

What are the main reasons you use AI tools frequently in your education or training?

Table showing the most common reasons learners frequently use AI tools in education or training, comparing 2024 and 2025. The most common uses include enhancing learning and understanding complex subjects, completing assignments more efficiently, preparing for exams, improving research and data analysis, and exploring additional learning resources. Usage levels are generally similar across the two years.

The impact of AI on learners’ classroom experiences and relationships is mixed, suggesting that the quality of implementation varies. 

  • 13% of learners responding to the 2025 survey said they feel more supported in their learning through AI‑assisted resources, a marked decrease from 29% in 2024.
  • AI is having mixed impacts on relationships in the classroom. The percentage of learners who reported spending more time collaborating with their peers as a result of AI (40%) is similar to the percentage who said they spend less time collaborating (41%). Likewise, 32% reported feeling more connected to their peers through enhanced group discussions and teamwork while 28% said they feel less connected.
  • 40% of learners reported increases in one-to-one time with teachers, while 30% reported decreases. 35% of learners said they feel more connected to their teachers, while 26% said they feel less connected.
  • In the 2025 survey, fewer learners overall reported that AI tools have had no significant impact on their relationships with either their teachers or instructors (12% vs. 23% in 2024) and their peers (11% vs. 32% in 2024).

In what way, if at all, are AI tools impacting your relationship with your teachers/instructors?

Stacked bar chart showing how AI tools impact learners’ relationships with teachers or instructors in 2024 and 2025. In 2024, 23% say AI makes them feel more connected to teachers and 18% say less connected. In 2025, 35% report feeling more connected and 26% less connected, indicating increases in both responses.

In what way, if at all, are AI tools impacting your relationship with your peers?

Bar chart showing the share of learners who say AI has had no significant impact on their relationship with peers. Thirty-two percent reported no impact in 2024, compared with 11% in 2025.

Methodology

AudienceNet surveyed 3,020 respondents aged 16 and above, between November 28 and December 8, 2025, boosting samples for JFF’s focus populations (all people without four-year college degrees, people of color and women whose highest level of education is a four-year degree, and people with records of arrest, conviction, or incarceration) to ensure robust insights into these populations. Final data were weighted to statistically reflect the U.S. population aged 16 and above, in accordance with the latest U.S. census data for comparison purposes. 

About the Center for Artificial Intelligence & the Future of Work

JFF’s Center for Artificial Intelligence & the Future of Work helps education and training institutions, employers, and public and private stakeholders better understand artificial intelligence and respond to AI’s impact on their workforce strategies. Contact us to learn more about how you can partner with us or support our work and subscribe for updates.