The Experience Advantage: Age-Inclusive AI Upskilling for the Longevity Workforce
Download the article here:
Listen to a review of this article:
Abstract: Workers aged 55 and older now make up nearly a quarter of the U.S. workforce, yet they are frequently left out of the AI capability-building reshaping organizations. This article synthesizes recent AARP survey data, early labor market evidence on generative AI, peer-reviewed research on age stereotypes, age-diversity climate, job design, and older adults' learning, and documented organizational cases. It finds that older workers are increasingly familiar with AI and interested in learning to use it, but that a large gap separates interest from training received, that age stereotypes about technology persist, and that AI-enabled hiring tools have created new legal exposure for age discrimination. Early evidence suggests that AI's employment effects have so far fallen most heavily on young entry-level workers rather than experienced ones. The article proposes five evidence-based responses: age-inclusive AI training design, bias-proofed hiring and AI tools, career conversations at every age, mutual mentoring in multigenerational teams, and flexible, phased work with structured knowledge transfer. It concludes with three long-term capabilities for sustaining an age-inclusive AI workforce.
Most conversations about artificial intelligence at work carry an unstated assumption about age. The people imagined as AI's natural adopters are young, digitally fluent, and early in their careers; the people imagined as its likely casualties are older, set in their ways, and close to the exit. The data now emerging from the United States tells a more complicated story, and in several respects the opposite one. Workers aged 55 and older made up 24% of the U.S. workforce in 2022, up from 10% in 1994, and they have been the fastest-growing age group in the labor force for more than two decades (Stinson & Wang, 2025). Early payroll evidence suggests that the employment effects of generative AI have so far fallen most heavily on workers in their early twenties in exposed occupations, while more experienced workers in the same occupations have held steady or grown (Brynjolfsson et al., 2025). And older workers are not refusing to learn. Nearly half of U.S. workers aged 50 and older say they want to learn more about using AI at work, yet only about one in eight has received any AI-related training (Perron, 2026c).
That combination produces a quiet but consequential mismatch. Organizations are adopting AI rapidly, and most say they are training their people to use it. Older employees, however, report being left out of that training, being stereotyped as less technologically capable, and facing hiring processes, increasingly mediated by algorithms, that may filter them out before a human ever reads their application. The legal system has begun to notice. In 2023, the U.S. Equal Employment Opportunity Commission (EEOC) settled what was widely described as its first case involving AI-enabled hiring discrimination, in which application software allegedly rejected older applicants automatically (U.S. Equal Employment Opportunity Commission, 2023). In 2025, a federal court allowed a nationwide age discrimination claim against the HR software provider Workday to proceed as a collective action (Holland & Knight, 2025).
This article argues that the longevity workforce, the growing population of people who work into their fifties, sixties, and beyond, is not a problem for AI strategy to manage but an asset that most AI strategies are failing to use. It reviews what is known about older workers' AI adoption, learning, and treatment; examines the organizational and individual consequences of getting this wrong; and sets out five evidence-based responses: age-inclusive AI training design, bias-proofing hiring and AI hiring tools, career conversations at every age, mutual mentoring within multigenerational teams, and flexible or phased work paired with deliberate knowledge transfer. It closes with three long-term capabilities that leaders can build so that the experience advantage compounds rather than erodes.
The Longevity Workforce and AI Landscape
Defining Key Terms
The phrase longevity workforce refers to the share of the labor force made up of people working later in life, usually defined in labor statistics as those aged 50 or 55 and older. It reflects longer lifespans, later retirement, and, for many, a financial or personal need to keep earning. Older worker is a moving category rather than a fixed one. U.S. federal law protects workers aged 40 and older under the Age Discrimination in Employment Act (ADEA), while most survey research, including AARP's, focuses on workers aged 50 and older. This article uses "older workers" to mean people aged 50 and older unless a source uses a different threshold, in which case the source's definition is noted.
Age inclusion describes the policies, practices, and climate through which an organization treats age as a dimension of diversity to be valued rather than a risk to be managed. Researchers often study it through the concept of age-diversity climate, defined as employees' shared perception that the organization treats workers of all ages fairly and values what each age group contributes (Boehm et al., 2014). Its opposite, an age discrimination climate, is the shared perception that age-based unfair treatment is common in the organization (Kunze et al., 2011).
AI upskilling refers to training that helps employees use AI tools effectively in their current roles, while reskilling refers to preparation for different roles. AI-enabled hiring tools are software systems that score, rank, sort, or screen job applicants using automated methods, ranging from simple knockout rules (such as age or date-of-birth filters) to machine-learning models that predict "fit." Finally, mutual mentoring, sometimes called reverse mentoring when a junior employee mentors a senior one, describes structured relationships in which learning flows in both directions across generations (Murphy, 2012).
Prevalence, Drivers, and Distribution
The demographic baseline is not in dispute. Beyond the national figures above, the U.S. Census Bureau has documented a structural shift in where older workers are employed. Between 2006 and 2022, employment at firms where at least a quarter of workers were 55 or older rose from about 13 million to about 35 million, while employment at firms with fewer than 10% older workers fell from about 45 million to about 32 million (Stinson & Wang, 2025). In utilities, the share of sector employment at firms with at least 25% older workers rose from 35% to 80% over that period. These are not niche employers. They are core parts of the industrial and service economy.
Older workers' labor market position is also distinctive. In September 2026, about 38 million Americans aged 55 and older were employed, and the unemployment rate for this group was 2.6%, compared with 4.2% for all workers (Schramm, 2026). Low unemployment, however, conceals difficulty once a job is lost: 28.1% of unemployed jobseekers aged 55 and older had been out of work for 27 weeks or more, and older jobseekers typically experience longer spells of unemployment than younger ones (Schramm, 2026). Field-experimental evidence helps explain why. In a resume correspondence study involving more than 40,000 applications, Neumark et al. (2019) found robust evidence of age discrimination in hiring against older women, particularly those near retirement age, with weaker evidence for men.
AI adoption among older workers is rising, but from a low base and with a stubborn training gap. AARP's three-wave survey of workers aged 50 and older, fielded through NORC's Foresight 50+ Omnibus, shows the share who use AI at work to a great or some extent rising from 14% in spring 2024 to 16% in early 2025 and 23% in March 2026 (Perron, 2026c). Familiarity with AI in the workplace rose from 39% to 52% over the same period. Interest in learning to use AI at work, however, has been essentially flat at 47% to 49%, while the share who have actually taken AI-for-work training or classes has crept from 7% to 12%. In the most recent wave, that left a 37-percentage-point gap between those who want training and those who have received it. Among employed respondents, 63% disagreed that their employer is doing enough to train workers on AI, and only 31% said their employer offers AI training to all workers regardless of age (Perron, 2026c).
The general workforce picture is consistent. A Pew Research Center survey from October 2024 found that 63% of U.S. workers did not use AI much or at all in their jobs, and that workers under 50 were somewhat more likely than those 50 and older to use it (17% versus 13%) (Lin & Parker, 2025). Notably, Pew found only modest age differences in whether workers had received AI-related training, which suggests that the problem for older workers is less a deep generational divide in learning than a failure of access and design in specific workplaces.
Employers tell a different story about the same reality. AARP's survey of 1,019 U.S. employers, fielded in November and December 2025, found that 88% say their organization uses AI and 83% provided AI-related training in the past year (Perron, 2026b). Almost all employers who offered training, 97%, said they take steps to make it accessible to employees of all ages. Yet high-engagement AI use, defined as three-quarters or more of employees using AI daily, was reported by 38% of employers overall but only 22% for employees aged 40 and older. Strikingly, 46% of employers said most or all decisions about how and where to use AI are made by employees aged 40 and older. In other words, older employees are often steering AI strategy while being under-served by AI training.
Finally, the distribution of AI's labor market effects so far appears to run against the stereotype. Using high-frequency payroll data from ADP, Brynjolfsson et al. (2025) found that, after generative AI became widely adopted, workers aged 22 to 25 in the most AI-exposed occupations experienced a 13% relative decline in employment, while employment for more experienced workers in the same occupations remained stable or continued to grow. A revised version of the analysis, using data through June 2026, reports that employment of 22- to 25-year-olds in exposed occupations is about 19% below where it would have been had it kept pace with less-exposed peers, that the gap stems mainly from reduced hiring rather than layoffs, and that experienced workers show no comparable gap (Brynjolfsson et al., 2026). The authors describe these as early descriptive indicators rather than causal estimates. A complementary analysis of LinkedIn data found that 49.4% of workers aged 50 and older hold roles considered insulated from generative AI disruption, compared with 42.2% of younger workers, largely because those roles rely on judgment, leadership, and collaboration (Perron, 2026d; Perron et al., 2025).
Organizational and Individual Consequences
Organizational Performance Impacts
The most immediate organizational cost of excluding older workers from AI capability-building is wasted human capital at precisely the moment it is most needed. Older employees hold the institutional context, client relationships, and judgment that make AI outputs usable. The AARP and LinkedIn analysis found that older professionals average about 15 more years of work experience and more than 10 more years in leadership roles than younger peers, and that their professional networks are 20.4% larger (Perron et al., 2025). When an organization deploys AI tools without equipping these employees to use them, it separates the tool from the people best placed to judge whether its outputs make sense.
Retention economics point the same way. In the same LinkedIn data, 85.4% of workers aged 50 and older hired in June 2024 were still with their employer a year later, compared with 70.6% of younger hires (Perron et al., 2025). Training investments in workers who stay deliver longer payback periods than the stereotype of the near-retiree suggests. Posthuma and Campion (2009), in their review of age stereotypes, noted that older workers are less likely to quit, that payback periods on training are often short, and that the common belief that older workers offer a poor return on training investment is weakly supported.
Climate matters for performance as well. In a study of 128 German companies and 8,651 employees, Kunze et al. (2011) found that organizational age diversity was associated with the emergence of an age discrimination climate, which in turn was associated with lower collective affective commitment and, through that, weaker firm performance. Age diversity on its own, in other words, does not guarantee benefits; when it is mismanaged, it can become a source of perceived unfairness that erodes commitment. Boehm et al. (2014), studying 93 small and medium-sized German firms and 14,260 employees, found the constructive counterpart: age-inclusive HR practices fostered a positive age-diversity climate, which was associated with stronger shared perceptions of social exchange and, indirectly, with higher firm performance and lower collective turnover intentions.
The macroeconomic stakes are large. Work by the Organisation for Economic Co-operation and Development (OECD) with AARP and the World Economic Forum estimated that building multigenerational workforces and expanding work opportunities for older employees could raise GDP per capita by almost 19% over three decades (Nazeri & Belelieu, 2020; Organisation for Economic Co-operation and Development, 2020). Meanwhile, employers responding to the World Economic Forum's Future of Jobs Report 2025 expect 39% of key skills to change by 2030 (Leopold, 2025). An organization that cannot reskill its experienced workers will be forced to replace them, often at higher cost and with lost know-how.
There is also a growing legal and reputational exposure. The iTutorGroup settlement and the Mobley v. Workday litigation, discussed below, show that automated screening tools can produce age discrimination claims against both employers and vendors. Employers that assume their AI hiring tools are neutral may discover otherwise in litigation, after the damage to applicants and to their employer brand has been done.
Individual Wellbeing and Stakeholder Impacts
For individuals, the consequences begin with the experience of age bias itself. AARP's 2025 survey of workers aged 50 and older found that 64% had seen or experienced age discrimination at work, unchanged from 2024, and 22% felt pushed out of a job because of their age (Perron, 2026a). The rates were higher among some groups: 71% of African American or Black older workers reported having seen or experienced age discrimination, and between 32% and 38% of African American or Black, Hispanic or Latino, and Asian American and Pacific Islander respondents felt pushed out. Subtle forms were common, and the single most frequently reported form was the assumption that older employees are less tech savvy (33%), followed by assumptions that they resist change (24%). One in five (20%) reported that younger employees were favored for training (Perron, 2026a).
These stereotypes are largely unsupported by evidence. In a meta-analysis covering hundreds of studies, Ng and Feldman (2012) examined six common stereotypes about older workers: that they are less motivated, less willing to participate in training and career development, more resistant to change, less trusting, less healthy, and more vulnerable to work-family imbalance. Only one, lower participation in training and career development, received empirical support, and even that finding is plausibly a result of how training opportunities are allocated rather than of older workers' preferences. An earlier meta-analysis by the same authors found that age was largely unrelated to core task performance and was positively related to several dimensions of organizational citizenship and safety behavior (Ng & Feldman, 2008). Stereotypes thus create a self-fulfilling loop: older workers are assumed to be less interested in learning, are offered fewer opportunities, participate less, and appear to confirm the stereotype.
AI introduces a specific anxiety into this loop. In AARP's March 2026 wave, 66% of older workers expected AI to replace workers and 63% expected it to displace them, while only 26% expected it to create new job opportunities (Perron, 2026c). Looking five years ahead, 30% saw AI mainly as a threat in their line of work and 19% mainly as an opportunity. Perron (2026d) reports that 24% of older workers believe AI could increase age bias in hiring, against 5% who think it could reduce it. These concerns are not irrational, given the lived experience of discrimination described above, and they are likely to shape whether older employees volunteer for AI pilots, disclose skills gaps, or seek new roles.
Other stakeholders are affected too. Younger employees lose access to mentors and to the tacit knowledge that never made it into documentation. Customers lose continuity of service and experienced judgment. The EEOC's statement on the iTutorGroup case made exactly this point: when hundreds of experienced tutors were screened out, students also lost access to them (U.S. Equal Employment Opportunity Commission, 2023). Finally, the early "canaries" evidence implies that entry-level pipelines may be narrowing as AI substitutes for junior tasks (Brynjolfsson et al., 2026). If organizations simultaneously push out experienced workers and hire fewer novices, they risk hollowing out the middle of their talent pipeline from both ends.
Evidence-Based Organizational Responses
The research and organizational experience reviewed above point toward five mutually reinforcing responses. None is exotic, and several draw on long-standing practice in age management. What is new is the urgency created by AI adoption and the legal scrutiny of AI-enabled hiring. Table 1 summarizes the responses.
Table 1: Evidence-Based Responses for an Age-Inclusive AI Workforce
Response | Evidence Base | Key Practices | Illustrative Examples |
Age-inclusive AI training design | Meta-analyses on older learners and training; AARP interest-versus-training gap | Universal eligibility, self-paced and task-based modules, active practice, peer cohorts | Barclays; Huntington Ingalls (Newport News Shipbuilding) |
Bias-proofing hiring and AI tools | Field experiments on age discrimination; EEOC enforcement; ADEA disparate-impact litigation | No birth-date collection pre-offer, age-impact audits, human review, vendor accountability | iTutorGroup (EEOC settlement); Workday (Mobley v. Workday) |
Career conversations at every age | Lifespan job design research; mid-life MOT pilots | Structured mid-career reviews covering work, skills, health, and finances | Aviva; Legal & General and Mercer pilots |
Mutual mentoring and multigenerational teams | Age-diversity climate research; reverse mentoring studies | Two-way mentoring, mixed-age AI pilot teams, shared credit | The Hartford; AARP Employer Alliance members |
Flexible, phased work and knowledge transfer | Ergonomic and job redesign evidence; age-inclusive HR practice research | Phased retirement, seasonal mobility, workstation redesign, structured handovers | BMW; Steelcase; CVS; PKF O'Connor Davies |
Each lever addresses one or more of the barriers documented above, and together they produce outcomes that no single program can deliver on its own: AI capability across all age groups, a positive age-diversity climate, retained judgment and know-how, and lower legal and reputational risk.
Age-Inclusive AI Training Design
The evidence on older adults' learning is clear on one point: older workers can and do learn, but how training is designed matters more for them than for younger learners. In a meta-analysis of training studies with older learners, Callahan et al. (2003) found that lecture, modeling, and active participation each contributed to training performance, and that self-pacing explained the largest share of variance among the instructional factors examined. Posthuma and Campion (2009) similarly concluded that while some laboratory studies find slower learning among older adults, field studies show smaller effects, and methods built on active participation and self-paced learning may be especially effective.
AARP's data adds an important nuance: older workers' motivation to learn is largely intrinsic. Earlier AARP research, summarized by Perron (2026d), found that 79% of workers aged 50 and older actively seek learning opportunities and that 86% would not accept a job without a chance to learn something new; personal interest, rather than employer mandate, was the dominant reason older workers gave for reskilling. Training that is framed as remedial ("catching up" on technology) is likely to feel demeaning. Training framed as an extension of expertise, applying AI to the work people already know well, is far more consistent with what motivates experienced learners.
The LinkedIn data suggest that when access is available, older professionals engage. The share of older workers' LinkedIn Learning sessions devoted to technology topics rose from 19.5% in 2022 to 26.6% in 2025, and the age gap in technology learning narrowed sharply over the same period (Perron, 2026d; Perron et al., 2025). Older workers increased their share of disruptive technology skills by 25% over five years, compared with 13.5% among younger workers (Perron, 2026d). The constraint, in short, is not willingness but opportunity.
Access also depends on formal eligibility rules, which vary by job level. Reporting on AARP's employer study, Perron (2026d) notes that 94% of employers make mid-level employees eligible for AI training, compared with 85% for senior-level employees and 77% for entry-level employees. Because older workers are distributed across all levels, and because experienced career changers may enter at lower levels, these eligibility rules can translate into age-related gaps even when no policy mentions age.
Effective approaches:
Universal eligibility: Make AI training available to all employees regardless of level, tenure, or proximity to retirement, and audit actual participation by age band, not just eligibility.
Task-anchored content: Build modules around real workflows employees already perform, so experienced workers apply AI to problems they understand deeply.
Self-paced and active formats: Offer self-paced modules, sandbox practice, and hands-on exercises rather than relying on one-off lectures, consistent with evidence on older learners.
Peer learning cohorts: Organize small mixed-age cohorts so participants learn from each other's questions and errors without public exposure.
Respectful framing: Present AI training as amplifying expertise and judgment, not as a remedial technology course.
Manager accountability: Track whether managers nominate older employees for AI pilots and advanced training, and make inclusive nomination part of leadership expectations.
Barclays launched its Bolder Apprenticeship programme in 2015, extending its apprenticeship scheme to people over 24 and describing itself as the first UK company to do so (HRreview, 2015). The bank's director of apprenticeships framed the decision in workforce terms, pointing to a projected UK skills shortage in which far more vacancies would need filling than there would be young people leaving education, and noting that older workers face a much harder task returning to employment once out of work. The programme's significance for AI upskilling is structural: it treats structured, earn-while-you-learn development as something an organization offers across the life course rather than only at the start of a career.
Huntington Ingalls Industries' Newport News Shipbuilding takes a similar view through its Apprentice School, which has no upper age limit, and through intergenerational mentoring in which younger workers help longer-tenured employees with new technology (Hannon, 2018). In a sector where tacit craft knowledge is critical and the workforce skews older, the shipbuilder pairs formal skill pathways open to all ages with informal two-way learning on the shop floor.
A common pitfall: Many organizations report that their AI training is age-inclusive because it is technically open to everyone. AARP's data show that the employer view (97% taking steps toward inclusive access) and the worker view (only 31% saying AI training is offered to all regardless of age) diverge sharply (Perron, 2026b, 2026c). Formal openness is not the same as practical access. If managers nominate younger employees for pilots, if training is scheduled around assumptions about who is "digital," or if older employees fear that asking for help will mark them as obsolete, participation will remain low regardless of policy.
Bias-Proofing Hiring and AI Hiring Tools
Age discrimination in hiring is well documented. Neumark et al. (2019) found robust evidence of discrimination against older women applicants in a large field experiment, and AARP's 2025 data show that job searches are the setting where older workers most often experience subtle age discrimination, cited by 37% of respondents (Perron, 2026a). AI-enabled hiring adds two new risks. First, automated tools can apply exclusionary rules at scale and without any human noticing. Second, machine-learning models trained on historical hiring decisions can reproduce the age patterns embedded in those decisions, producing what U.S. law calls disparate impact: a neutral-seeming practice that disproportionately disadvantages a protected group. Under the ADEA, workers aged 40 and older are protected, and disparate-impact claims do not require proof of intent.
Practitioners advising employers after the Workday ruling have converged on a set of controls: auditing vendors and requiring documentation of bias testing, preserving meaningful human oversight of automated recommendations, documenting hiring criteria, monitoring outcomes by age and other protected characteristics, and establishing formal governance over AI hiring tools (Fisher Phillips, 2025). Older workers' own perceptions underline the urgency; as noted earlier, far more believe AI will increase age bias in hiring than reduce it (Perron, 2026d).
Effective approaches:
Remove age proxies: Do not collect birth dates or graduation years before an offer, and review knockout questions and filters for rules that effectively screen by age.
Audit for age impact: Test AI screening and ranking tools for differential pass-through rates for applicants aged 40 and older, both before deployment and periodically after.
Contract for accountability: Require vendors to document bias testing, share relevant data, and support adverse-impact analysis, and include nondiscrimination terms in contracts.
Keep humans meaningfully in the loop: Train recruiters to question and override automated rankings and to review a sample of rejected applicants.
Use skills-based criteria: Replace "years of experience" ranges and "digital native" language with demonstrable skill requirements.
Monitor and govern: Assign clear ownership for AI hiring governance and report outcomes to senior leadership.
iTutorGroup, an online English-language tutoring business, paid $365,000 in 2023 to settle an EEOC lawsuit alleging that its tutor application software automatically rejected female applicants aged 55 and older and male applicants aged 60 and older (U.S. Equal Employment Opportunity Commission, 2023). More than 200 qualified U.S.-based applicants were allegedly rejected because of age. According to legal commentary on the filing, the charging party was rejected when she applied with her real birth date and was offered an interview the next day after submitting an otherwise identical application with a more recent birth date (Baker Donelson, 2023). The consent decree barred the company from requesting birth dates before a job offer and required anti-discrimination policies, training for hiring staff, EEOC monitoring for at least five years, and outreach to rejected applicants if the company resumes U.S. hiring. The company denied the allegations. The case shows how a crude, rule-based filter embedded in software can do at scale what an individual recruiter could not easily hide.
Workday, the HR software provider, faces a more complex test of vendor responsibility in Mobley v. Workday. Derek Mobley, an applicant over 40, alleges he applied for more than 100 positions at companies using Workday's platform and was rejected every time, and that Workday's AI tools for scoring, ranking, and screening applicants disproportionately disadvantage older jobseekers (Holland & Knight, 2025). In May 2025, Judge Rita Lin of the U.S. District Court for the Northern District of California granted preliminary certification of a nationwide collective action under the ADEA covering applicants aged 40 and older who were denied employment recommendations through Workday's platform since September 2020 (Fisher Phillips, 2025; Holland & Knight, 2025). Court records indicate that the court later addressed whether applicants processed with Workday's HiredScore AI features fall within the collective, approved a notice plan in late 2025, and continued to rule on motions in 2026 (Civil Rights Litigation Clearinghouse, 2025). Workday contends the claims lack merit, and no finding of liability has been made. For employers, the case signals that delegating screening to a vendor does not delegate away the question of disparate impact.
A common pitfall: Organizations often treat bias auditing as a one-time procurement check focused on race and sex. Age can be overlooked entirely, both because some audit regimes do not require it and because age proxies, such as graduation years, years-of-experience caps, or "recent graduate" preferences, are easy to miss. A tool that passes a narrow audit at purchase can still drift as it learns from new data or is configured differently by local recruiters.
Career Conversations at Every Age
Career development in many organizations is front-loaded. Early-career employees receive structured development plans, while employees in their fifties are presumed to be on a glide path. Lifespan approaches to work challenge that assumption. Truxillo et al. (2012) argue that the role of age in job design has been largely neglected, and that the fit between job characteristics and workers' changing needs, motives, and capabilities across the lifespan shapes satisfaction, engagement, and performance. Truxillo et al. (2015), reviewing workplace interventions for an aging workforce, call for practices that help workers and organizations adjust jobs over time rather than assume a fixed fit.
The mid-life MOT (named after the UK's annual vehicle roadworthiness test) is one practical expression of this idea: a structured review, typically offered in people's forties, fifties, and sixties, that covers work and skills, health and wellbeing, and finances together. A joint report by the Centre for Ageing Better, drawing on pilots at Aviva, Legal & General, the Pensions Advisory Service, and Mercer, found that participants valued combined guidance across these domains and that the reviews worked best as an ongoing process with practical outputs and follow-up rather than a one-off event (Centre for Ageing Better, 2018). In the AI context, a mid-career checkup is also the natural place to discuss how AI is changing a person's role, what new skills they want to build, and how their expertise could be redeployed.
Effective approaches:
Routine, not remedial: Offer structured career conversations at regular intervals for everyone, including employees in their fifties and sixties, so they are not perceived as a signal of exit.
Integrated scope: Combine discussion of skills and AI readiness with health, caregiving, flexibility, and financial planning, which shape decisions about how long and how to work.
AI role mapping: Use the conversation to identify which tasks in a role AI is likely to change and what new or adjacent roles might suit the person's experience.
Clear outputs: End with a concrete development plan, training commitments, and a follow-up date.
Trained facilitators: Prepare managers or coaches to conduct age-neutral conversations, avoiding assumptions about retirement plans.
Aviva launched a Mid-Life MOT for UK employees aged 45 and older in 2018, after piloting it with 100 employees aged 45 to 60 (Aviva, 2018). The program addressed three areas: wealth, through planning tools and access to financial advice; work, through mid-career reviews, learning and development support, apprenticeship information for all ages, smart working, and phased retirement options; and wellbeing, through health support and an expanded policy for carers. Aviva described its goals as helping older employees review, reposition, and retain, and noted that its oldest apprentice at the time was 68.
Legal & General and Mercer were among the organizations whose mid-life review pilots informed the Centre for Ageing Better's 2018 guidance, alongside Aviva and the Pensions Advisory Service, which focused on self-employed people (Centre for Ageing Better, 2018). The pilots experimented with telephone, face-to-face, group, and online delivery, and the shared lesson was that focus and follow-through mattered more than format.
A common pitfall: Career conversations with older employees can quickly become retirement conversations, which raises legal risk and signals that the organization has already written the person off. A conversation that opens with "When are you planning to retire?" forecloses the discussion that matters most: how the employee's expertise can be extended, updated, and redeployed as AI changes the work.
Mutual Mentoring and Multigenerational Teams
Employers see value in age diversity. In AARP's 2025 employer survey, two-thirds (66%) said having employees from multiple generations is very important, and large majorities said mixed-age teams boost productivity (93%), increase empathy across generations (88%), and strengthen informal mentoring (87%) (Perron, 2026b). Research on age-diversity climate explains why these benefits are not automatic: diversity helps when the climate signals that all age groups are valued, and hurts when it breeds perceived discrimination (Boehm et al., 2014; Kunze et al., 2011).
Mutual mentoring is one of the most direct ways to build that climate around AI. In classic reverse mentoring, younger employees help senior colleagues learn new technologies, while senior colleagues provide perspective on organizational context, judgment, and career navigation (Murphy, 2012). The AI era makes the two-way design more important, not less. Younger employees may be fluent with new tools but lack the domain knowledge needed to tell when an AI output is wrong; experienced employees can supply that judgment. AARP's press materials note that 40% of younger workers say older colleagues help them stay current with tech trends, which suggests that technology learning already flows in both directions (AARP, 2026).
Effective approaches:
Design for reciprocity: Frame pairings as mutual learning, with explicit goals for both partners, rather than as one-way "digital tutoring" of senior staff.
Mixed-age AI pilots: Staff AI experimentation teams deliberately across age groups so that tool fluency and domain judgment are combined from the start.
Psychological safety: Establish norms that make it acceptable for senior people to ask basic questions and for junior people to challenge outputs.
Shared credit: Recognize both mentors and mentees, and capture the process and workflow changes that result.
Climate measurement: Include age-diversity climate items in engagement surveys to detect whether mixed-age teams are working.
The Hartford, the insurance company, introduced reverse mentoring partnerships under then-CEO Liam McGee so executives could become more fluent in the digital technologies used by customers and partners (Ellis, 2013). The program began with a small group of millennial employees mentoring managers, focusing on social media, and grew to more than 50 mentees across seven states. It was credited with business changes including new telemarketing tactics, an updated social media policy, and greater comfort with information sharing on the company network. In the first wave, 11 of 12 mentors were promoted within a year (Ellis, 2013), an indication that well-designed reverse mentoring develops the younger partner as well as the senior one.
The AARP Employer Alliance, launched in March 2026, offers a different model: a community of employers, including organizations such as the American Heart Association and Sanofi, that share practices for building multigenerational workplaces (AARP, 2026). The alliance has two tiers, one for organizations evaluating their current practices and one for those ready to implement, or already implementing, age-friendly policies. Its premise is that multigenerational teams are a source of innovation and that employers learn faster from one another than alone.
A common pitfall: Reverse mentoring programs can reinforce the very stereotype they aim to overcome if they are framed as young people teaching older people "how technology works." When senior employees are cast only as learners, and junior employees only as tech helpers, both groups lose. The most effective programs make the exchange explicit in both directions and connect it to real business problems
Flexible, Phased Work and Knowledge Transfer
Many older workers want to keep working, but not necessarily in the same way. As Nazeri and Belelieu (2020) summarized in reporting on the OECD's collaboration with AARP and the World Economic Forum, 57% of workers globally want to keep working past traditional retirement ages, but fewer than one in three can move from full-time to part-time work. This mismatch drives a costly binary: employees either keep a full-time role that no longer fits or leave entirely, taking their knowledge with them. Job design research suggests a better alternative. Truxillo et al. (2012) argue that job characteristics such as autonomy, task variety, and physical demands interact with age, so redesigning work can sustain engagement and performance as needs change. Boehm et al. (2014) found that age-inclusive HR practices, including flexible arrangements, contribute to the climate that drives firm-level outcomes.
AI changes the knowledge transfer problem in two ways. It can help capture and organize expertise, but it also raises the value of tacit judgment that does not easily become data. Phased retirement and flexible roles create time for experienced employees to coach successors, validate AI-assisted workflows, and document the reasoning behind decisions.
Effective approaches:
Phased retirement options: Offer reduced hours or part-year arrangements with continued benefits, rather than forcing a choice between full-time work and departure.
Knowledge transfer roles: Formally assign phased retirees to coach, review, and document, with protected time and clear deliverables.
Ergonomic and process redesign: Involve older workers in redesigning workstations and processes, which often benefits all ages.
Mobility and flexibility: Offer location flexibility, seasonal mobility, and part-time or consulting arrangements.
AI-assisted capture: Use AI tools to organize documented expertise, with experienced staff validating what is captured.
BMW offers the best-known example of age-inclusive job redesign. In a pilot at its plant in Dingolfing, Germany, the company staffed an assembly line to reflect the average age of 47 that it projected for its workforce in 2017, and worked with line employees in workshops to identify improvements (Loch et al., 2010). The team implemented about 70 changes to equipment, workstations, and work practices at a direct cost of about €20,000. Productivity on the line rose 7% in one year, absenteeism fell from 7% to 2%, below the plant average, and the line met its defect target. Most of the changes were modest and inexpensive, which is precisely the point: redesign driven by the workers themselves made the line more productive for everyone on it.
Steelcase, the office furniture maker, began a phased retirement plan in 2012 that allows employees to reduce their hours rather than leave, in response to the coming retirement of many baby boomers in information technology and skilled manufacturing trades (Hannon, 2018). One participant profiled by Hannon was able to keep his income and health coverage while transitioning toward retirement, and the company gains time for experienced workers to pass on skills that are in short supply.
CVS, the pharmacy retailer, was recognized in 2007 with the American Society on Aging's Business and Aging Award for recruiting, training, and retaining older workers and for programs that passed their knowledge to younger employees (Drug Topics, 2007). Its Snowbird program allowed employees to transfer seasonally to stores in other locations, filling roles such as pharmacist and photo supervisor. At the time, the share of CVS employees aged 50 and older had risen from 7% in the early 1990s to more than 17%. In professional services, the accounting network PKF O'Connor Davies has hired partners over 60 and offers shorter workweeks, flexible hours, part-time and remote options, and consulting arrangements for employees nearing retirement (Hannon, 2018).
A common pitfall: Flexible and phased arrangements often exist only informally, negotiated case by case by employees with strong relationships or rare skills. This produces inequity and makes knowledge transfer accidental rather than planned. Without formal eligibility criteria, documented expectations for knowledge sharing, and benefit continuity, phased retirement tends to benefit a few favored individuals while the broader workforce continues to face an all-or-nothing choice.
Building Long-Term Longevity Workforce Capability
The five responses above can be launched within a year. Sustaining an age-inclusive AI workforce, however, requires durable capabilities that outlast individual programs and leaders. Three stand out.
Age-Inclusive AI Governance and Measurement
Most organizations now have some form of AI governance, typically focused on data privacy, security, and accuracy. Few treat age as an explicit dimension of that governance. Building long-term capability means extending AI governance to cover who gets trained, who gets selected for AI-enabled roles, and how AI hiring and performance tools affect workers of different ages. AARP's finding that nearly half of employers say employees aged 40 and older make most strategic AI decisions (Perron, 2026b) suggests that older employees are already central to AI governance; the task is to make sure their own treatment falls within its scope.
In practice, this means building a small set of age-disaggregated metrics into regular workforce reporting: AI training participation and completion by age band, AI tool usage, selection rates for AI-related projects and promotions, hiring pass-through rates for applicants aged 40 and older, and voluntary exits by age. It means requiring that every AI tool used in hiring, promotion, or performance management undergo age-impact testing before deployment and at regular intervals. And it means giving a named executive accountability for these results, with reporting to the board or its equivalent. Given ongoing litigation such as Mobley v. Workday, this is not only good practice but prudent risk management.
Lifelong Learning Infrastructure
The second capability is a learning system designed for careers that may span 50 years. The World Economic Forum's estimate that 39% of key skills will change by 2030 (Leopold, 2025) implies that every employee, at every age, will need to retool repeatedly. Organizations that treat learning as an early-career investment will find themselves with a workforce whose most experienced members are least equipped for the technologies reshaping their work.
Lifelong learning infrastructure includes learning budgets and time allowances that do not decline with age or tenure; modular, self-paced AI curricula that experienced workers can tailor to their roles; apprenticeship and returnship pathways open to people at any career stage, as Barclays and Newport News Shipbuilding have demonstrated; and regular career conversations that connect learning to concrete opportunities. It also includes learning from the people being trained. Older workers' interest in AI training has been stable at nearly half for three years (Perron, 2026c); that demand is a resource. Organizations can ask experienced employees to co-design training, identify which tasks AI should and should not take over, and serve as internal faculty for domain-specific AI applications.
A Culture of Generational Reciprocity
The third capability is cultural. Age-diversity climate research shows that the same demographic mix can strengthen or weaken performance depending on whether employees believe all ages are valued (Boehm et al., 2014; Kunze et al., 2011). In the AI era, a culture of generational reciprocity is one in which younger employees' fluency with new tools and older employees' judgment, networks, and institutional knowledge are seen as complementary, and in which both groups have reason to share what they know.
Such a culture is built through accumulated signals: leaders who model learning AI tools publicly regardless of age; recognition systems that reward knowledge sharing and mentoring; language in job postings and internal communications that avoids age-coded terms; and stories that celebrate late-career growth as well as early-career achievement. It is also reinforced by the evidence that entry-level workers, not experienced ones, are currently bearing the brunt of AI-related hiring slowdowns (Brynjolfsson et al., 2026). That finding gives organizations a shared interest across generations. Experienced employees are well placed to help younger colleagues develop the judgment that makes them valuable alongside AI, and younger employees are well placed to help experienced colleagues extend their expertise through new tools. A culture that sees these exchanges as mutual investment will be more resilient than one that pits generations against each other.
Conclusion
The longevity workforce and the AI transition are colliding in most organizations, often without anyone noticing. Workers aged 55 and older are now nearly a quarter of the U.S. workforce, and they are concentrated in sectors central to the economy. They are increasingly familiar with AI, increasingly using it, and consistently interested in learning more. Yet only a small minority have been trained, many experience age stereotyping (most commonly the assumption that they are less tech savvy), and they face hiring processes in which automated tools may reproduce or amplify age bias. Meanwhile, the early evidence on AI's labor market effects suggests that experienced workers in exposed occupations have so far held their ground better than young entrants, and that a large share of older workers hold roles that rely on judgment, leadership, and collaboration.
The organizational case for age inclusion is therefore stronger, not weaker, in the age of AI. Experienced workers provide the context and judgment that make AI outputs trustworthy. They stay longer, so training investments pay back. They are often already shaping AI strategy. And excluding them, whether through training gaps, stereotyped assumptions, or biased tools, carries rising legal, reputational, and performance costs. The research on age-diversity climate shows that these effects compound: inclusive practices build a climate that sustains commitment and performance, while discriminatory climates erode both.
Leaders do not need new theories to act. They need to apply well-established evidence about how experienced adults learn, how bias enters selection, how jobs can fit changing lives, and how knowledge moves between generations, and to apply it deliberately to AI adoption.
Close the training gap: Make AI training universally available, self-paced, and task-anchored, and track participation by age, not just eligibility.
Bias-proof hiring: Remove age proxies, audit AI hiring tools for age impact, retain meaningful human oversight, and hold vendors accountable, mindful of iTutorGroup and Mobley v. Workday.
Talk careers at every age: Offer structured mid-career checkups that integrate AI readiness, skills, health, and finances, and keep retirement assumptions out of them.
Make mentoring mutual: Pair tool fluency with domain judgment in mixed-age AI pilots and two-way mentoring.
Design flexible exits and handovers: Formalize phased retirement and flexible work, and build knowledge transfer into them.
Govern and measure: Add age to AI governance, report age-disaggregated workforce metrics, and assign executive accountability.
Invest for the long run: Build lifelong learning infrastructure and a culture of generational reciprocity so that experience becomes a compounding advantage.
Research Infographic

References
AARP. (2026, March 9). AARP launches Employer Alliance to help organizations unlock the value of multigenerational teams [Press release].
Aviva. (2018, October 18). Aviva launches "Mid-Life MOT" for its over 45s [Press release].
Baker Donelson. (2023). The EEOC settles its first lawsuit alleging AI-based discrimination in employment.
Boehm, S. A., Kunze, F., & Bruch, H. (2014). Spotlight on age-diversity climate: The impact of age-inclusive HR practices on firm-level outcomes. Personnel Psychology, 67(3), 667–704.
Brynjolfsson, E., Chandar, B., & Chen, R. (2025). Canaries in the coal mine? Six facts about the recent employment effects of artificial intelligence [Working paper]. Stanford Digital Economy Lab.
Brynjolfsson, E., Chandar, B., & Chen, R. (2026). Canaries in the coal mine? Six facts about the recent employment effects of artificial intelligence (Rev. ed., August 12, 2026). Stanford Digital Economy Lab.
Callahan, J. S., Kiker, D. S., & Cross, T. (2003). Does method matter? A meta-analysis of the effects of training method on older learner training performance. Journal of Management, 29(5), 663–680.
Centre for Ageing Better. (2018, September 25). Industry heavyweights lead development of "mid-life MOT".
Civil Rights Litigation Clearinghouse. (2025). Mobley v. Workday, Inc. (Case No. 3:23-cv-00770, N.D. Cal.) [Case record]. University of Michigan Law School.
Drug Topics. (2007, March 19). CVS receives award for mature worker initiatives. Drug Topics.
Ellis, R. K. (2013, September). Reverse mentoring: Letting millennials lead the way. TD Magazine. Association for Talent Development.
Fisher Phillips. (2025, May 20). Discrimination lawsuit over Workday's AI hiring tools can proceed as class action: 6 things employers should do.
Hannon, K. (2018, March 2). Reaping the benefits of an aging workforce. The New York Times.
Holland & Knight. (2025, May 27). Federal court allows collective action lawsuit over alleged AI hiring bias.
HRreview. (2015, September 4). Barclays launches apprenticeship programme for older workers. HRreview.
Kunze, F., Boehm, S. A., & Bruch, H. (2011). Age diversity, age discrimination climate and performance consequences: A cross organizational study. Journal of Organizational Behavior, 32(2), 264–290.
Leopold, T. (2025, January 8). Future of Jobs Report 2025: The jobs of the future and the skills you need to get them. World Economic Forum.
Lin, L., & Parker, K. (2025, February 25). Workers' exposure to AI. Pew Research Center.
Loch, C. H., Sting, F. J., Bauer, N., & Mauermann, H. (2010). How BMW is defusing the demographic time bomb. Harvard Business Review, 88(3), 99–102.
Murphy, W. M. (2012). Reverse mentoring at work: Fostering cross-generational learning and developing millennial leaders. Human Resource Management, 51(4), 549–574.
Nazeri, H., & Belelieu, A. (2020, December 16). How a multi-generational workforce is key to economic growth. World Economic Forum.
Neumark, D., Burn, I., & Button, P. (2019). Is it harder for older workers to find jobs? New and improved evidence from a field experiment. Journal of Political Economy, 127(2), 922–970.
Ng, T. W. H., & Feldman, D. C. (2008). The relationship of age to ten dimensions of job performance. Journal of Applied Psychology, 93(2), 392–423.
Ng, T. W. H., & Feldman, D. C. (2012). Evaluating six common stereotypes about older workers with meta-analytical data. Personnel Psychology, 65(4), 821–858.
Organisation for Economic Co-operation and Development. (2020). Promoting an age-inclusive workforce: Living, learning and earning longer. OECD Publishing.
Perron, R. (2026a). Age discrimination holds steady among older workers in 2025. AARP Research.
Perron, R. (2026b). AI, employers, and the multigenerational workforce: How U.S. employers are using artificial intelligence. AARP Research.
Perron, R. (2026c). AI training lags behind for older workers even with increasing familiarity. AARP Research.
Perron, R. (2026d, June 24). Not left behind: Older workers, artificial intelligence, and the data behind adoption and adaptation. Generations Journal. American Society on Aging.
Perron, R., Baird, M., & Hood, R. (2025). The untapped value older workers bring to the multigenerational workforce. AARP Research.
Posthuma, R. A., & Campion, M. A. (2009). Age stereotypes in the workplace: Common stereotypes, moderators, and future research directions. Journal of Management, 35(1), 158–188.
Schramm, J. (2026, October 2). September 2026 employment data digest. AARP Public Policy Institute.
Stinson, M., & Wang, S. (2025, December 2). Firms in production sectors and northern states have some of the highest shares of older workers. U.S. Census Bureau.
Truxillo, D. M., Cadiz, D. M., & Hammer, L. B. (2015). Supporting the aging workforce: A review and recommendations for workplace intervention research. Annual Review of Organizational Psychology and Organizational Behavior, 2, 351–381.
Truxillo, D. M., Cadiz, D. M., Rineer, J. R., Zaniboni, S., & Fraccaroli, F. (2012). A lifespan perspective on job design: Fitting the job and the worker to promote job satisfaction, engagement, and performance. Organizational Psychology Review, 2(4), 340–360.
U.S. Equal Employment Opportunity Commission. (2023, September 11). iTutorGroup to pay $365,000 to settle EEOC discriminatory hiring suit [Press release].

Jonathan H. Westover, PhD, Chief Research Officer (Nexus Institute for Work and AI); Co-Founder & Chief Workforce and Learning Officer (Future State University); Founder & CEO (Human Capital Innovations); Professor of Organizational Leadership & Change (UVU). Read Jonathan Westover's executive profile here.
Suggested Citation: Westover, J. H. (2026). The Experience Advantage: Age-Inclusive AI Upskilling for the Longevity Workforce. Human Capital Leadership Review, 40(4). doi.org/10.70175/hclreview.2020.40.4.3






















