The rapid advancement of artificial intelligence (AI), and especially generative AI, is creating new opportunities and challenges for workforce systems and training organizations. Through listening sessions over the course of 2025, we engaged with nearly 20 representatives of workforce, training, educational, and other community-based organizations that work with lower-income populations to help them prepare for, attain, and advance in employment opportunities. These organizations are grappling with how to best support their clients in adapting to these emerging tools and the evolving nature of work. Our roundtable and listening session conversations yielded the following insights from these stakeholders.
Listening session participants had not yet seen an increase in employer demand for AI-related skills but still saw value in their clients understanding how to work with AI.
The stakeholders we spoke with shared that they had yet to see significant shifts in the kinds of skills in demand from local employers or in job descriptions being posted for recruitment. That aligns with recent research from the Federal Reserve Bank of Atlanta, which found that, among job postings requiring less than a bachelor’s degree, less than 2% of openings required a specific AI skill.
At the same time, respondents recognized the rapid pace of change and increasing adoption of AI. One participant predicted, “There will be a line [crossed] where most people are using [AI].” Another noted, “People not exposed [to AI] now are going to fall further and further behind…We have low-income people not using it who will fall further behind because they don’t know how.” What was less clear to participants was how to best integrate AI training into current workforce training systems and supports.
Respondents were still weighing how best to target their training and program resources as their organizations and clients navigate the AI transition.
The organizations we spoke with were grappling with questions that many characterized as fundamental or ethical as they considered how their programming might need to shift in the era of AI. On one hand, respondents were concerned about training their clients for jobs that may have once seemed like good opportunities but may be at risk of being replaced or restructured by AI, including office and administrative support roles that now rank among the most exposed to AI. One respondent shared their concerns about “investing in [training for] jobs that won’t be there in two years. Is that a good investment if that skill is no longer needed in the workforce?” That is “coming up a lot” in their organization’s strategic conversations.
Others wondered how much they should focus their resources on training for occupations that are more likely to be “safe” from AI, such as skilled trades that require physical tasks. Yet those same respondents acknowledged that those jobs may not remain “safe” as technology—including both AI and robotics—evolves. They also didn’t want their clients to be insulated from AI to the extent that they couldn’t be competitive in the labor market if they needed or wanted to transition away from less AI-exposed occupations.
In terms of finding the best ways to equip their clients and students with AI skills, some respondents have been exploring smaller pilot efforts—in close alignment with local employers—to create AI credentials or short-term training. However, most respondents we spoke with were not currently planning to develop specific credentials—noting how quickly AI tools and adoption are evolving—but instead were focusing on ways to integrate AI tools in their trainings and programs. For instance, a community college representative shared how they were finding ways to incorporate AI tools into course curricula across several different training pathways, not as the focus of the course but as tools to help complete the coursework and allow students to become familiar with how they work.
Discussants identified challenges—including the existing digital divide—that might contribute to gaps in AI skill building.
Even with the challenges of adapting programming during a dynamic period of technological change, listening session participants expressed hopes that future AI tools could support providers in further tailoring education and upskilling to each individual, helping to create or bridge career pathways.
However, participants also voiced concerns about potential downside risks of this AI transition. Among their concerns, participants noted the existing digital divide could create additional barriers to building AI skills for those with digital literacy gaps or without reliable internet access. Respondents also said that many younger lower-income people just entering the workforce tend to access the internet through their mobile devices and don’t have access to desktop or laptop computers, which can make it more difficult to learn how to use these tools in a work context.
Stakeholders also expressed concerns for older workers who may struggle to adapt to this new technology. At the same time, participants discussed age as potentially both a risk factor and protective: although older workers are not “AI-native” (or, in many cases, computer-native), they possess many of the real-world critical thinking and conflict resolution skills that come from years in the workforce. While accumulated experience may help buffer older workers, it is less clear how the next generation will develop the experience and informed expertise essential for their own career advancement if AI adoption erodes entry-level positions.
Participants reflected on the importance of fostering and honing critical thinking skills to work effectively alongside AI, including understanding its capabilities, limitations, and appropriate applications.
Participants posited that workers would need to figure out how to adapt alongside evolving AI tools to be best positioned in the labor market moving forward. Respondents noted that putting the primary emphasis on helping clients and students develop critical-thinking and other essential “human,” or soft, skills would both help them understand how to effectively use AI to augment their work, and help them develop skills that have long been among the most requested by employers. And in the case of younger workers whose entry-level or early career paths may look different given the advent of AI, participants noted that it will be especially important to emphasize and support the development of critical-thinking and soft skills for these workers. Respondents observed that, at least so far, soft skills and emotional intelligence are not replicable by AI, making this a fundamentally human comparative advantage across many occupations and tasks.
As the AI landscape evolves, workforce training organizations and educational institutions face the challenge of preparing workers for a future that’s still taking shape. Although the full impact of AI on employment remains to be seen, it is clear that developing adaptable, critical thinking skills and a culture of continuous learning will be important for workers to thrive in an AI-augmented workplace. The coming years will likely see significant changes in how education and workforce training providers meet these emerging needs.
This article is part of a Community Investments series exploring the ways in which the growing prevalence of artificial intelligence may be impacting economic conditions, especially in low- and moderate-income communities and among community development stakeholders. Gaining greater insight into emerging economic trends through community engagement and analysis—including better understanding the economic experiences of lower-income workers and consumers—contributes to the Federal Reserve Bank of San Francisco’s work to support monetary policy, strengthen financial institutions, and enhance the payments systems.