Artificial intelligence adoption across American workplaces has quadrupled in just two years, yet employer-provided training programmes have failed to keep pace with the rapid transformation, according to a national survey released in March 2025 by Jobs for the Future.
The study found that 35% of American workers now use AI regularly at work, a dramatic increase from just 8% in 2023. However, only 31% reported receiving AI training from their employers, revealing a widening skills gap that threatens to leave millions of workers unprepared for technology-driven workplace changes.
More than half of workers – 56% – said they do not feel adequately prepared to use AI in their current roles, despite widespread adoption of the technology across industries. The disconnect highlights a critical challenge for HR professionals and business leaders as artificial intelligence becomes embedded in daily operations.
Self-Directed Learning Dominates
The Jobs for the Future survey, conducted by AudienceNet between 20 and 27 November 2024 with 2,754 respondents, found that workers are largely taking matters into their own hands. Sixty percent reported using AI primarily for self-directed learning rather than through structured employer programmes.
“With AI already transforming the future of work and learning, access to training, tools, and the opportunity to help shape this technology are more critical than ever,” said Kristina Francis, executive director of JFFLabs. “We risk widening divides if we don’t act now.”
Learners currently enrolled in education or training showed even higher adoption rates, with 59% reporting at least weekly AI use. Yet the skills infrastructure supporting this transition remains underdeveloped, particularly for workers without four-year degrees, people of colour, and those with criminal records.
Unequal Impact Across Demographics
The survey revealed significant disparities in how AI’s impact is felt across different populations. Whilst 53% of all respondents felt they needed to gain new skills due to AI in the next five years, this concern was more acute amongst respondents of colour (70%) and those with criminal records (56%).
Nineteen percent of all respondents said they have already used AI tools to obtain better employment, whilst 30% of respondents of colour indicated they are either actively pursuing different careers or considering career changes due to AI-driven transformation.
“AI should make us all better off by creating quality jobs, pathways to entrepreneurship, sustainable livelihoods, and opportunities to unleash human agency,” Francis said. The survey suggests, however, that without deliberate intervention, the technology may exacerbate existing inequalities rather than reduce them.
Market Realities Versus Aspirations
Recent analysis of AI career trends reveals a complex landscape where job seeker interest, market accessibility, and compensation remain misaligned. Entry-level positions face particular pressure, with research suggesting young workers in AI-exposed sectors experienced employment declines whilst older workers’ employment grew.
PwC’s 2025 Global AI Jobs Barometer, which analysed close to one billion job advertisements from six continents, found that AI-exposed industries have seen revenue growth accelerate sharply since 2022. The research suggested that AI can make workers more valuable rather than redundant, even in highly automatable roles.
The World Economic Forum’s Future of Jobs Report 2025 projects that technology will be the most disruptive force in labour markets over the next five years. Advances in AI and information-processing technologies are expected to create 19 million jobs whilst displacing nine million others, with skill gaps cited as the leading barrier to digital transformation by 63% of employers surveyed.
Training Infrastructure Inadequate
The challenge extends beyond individual employers. JFF’s review of leading AI-literacy curricula found many are written at a grade 11 reading level, well above the proficiency of millions of adult learners who need these skills most urgently.
Michael Collins, senior vice president of population strategies at JFF, emphasised the dual nature of the challenge: “AI is a powerful tool that creates efficiencies, but we also want to think about how we can augment what is uniquely human in a way that improves job quality. The challenge, and the opportunity here, is to support learners and workers to take advantage of AI technology.”
The survey found that 77% of respondents believe AI will impact their job or career in the next three to five years, indicating widespread recognition that change is coming. Yet 88% of workers do not trust their employers to support them in understanding AI, according to earlier JFF research.
Call for Systemic Response
The findings underscore the urgent need for coordinated action amongst employers, education providers, and policymakers. JFF has developed a call to action urging stakeholders to ensure AI improves job quality, fosters entrepreneurship, and promotes economic advancement rather than concentrating benefits amongst those already advantaged.
For HR leaders and senior executives, the research suggests three immediate priorities: developing comprehensive AI training programmes accessible to all employees, particularly those without advanced degrees; investing in uniquely human skills such as critical thinking, communication, and adaptability that complement rather than compete with AI capabilities; and ensuring training resources meet workers where they are in terms of literacy and technical proficiency.
The gap between AI adoption and adequate preparation represents both a risk and an opportunity. Organisations that move quickly to address the training deficit may gain competitive advantage through a more capable, confident workforce. Those that delay risk being overtaken by more agile competitors whilst simultaneously facing increased employee attrition and diminished organisational capacity.
As AI continues its rapid integration into workplace operations across all sectors, the question is no longer whether workers need new skills, but whether employers and education systems can deliver them fast enough to prevent widespread disruption.
