Generative AI is rapidly reshaping the world of work, raising a fundamental question for European skills policy: will it widen inequality or help reduce it? The answer depends less on the technology itself and more on how it is governed, accessed, and embedded in education and training systems, including vocational education and training (VET).
On one side of the debate, automation presents clear risks. Generative AI can replace routine cognitive tasks in roles such as administrative assistance, data entry, basic accounting, or customer support. These occupations often provide entry-level employment and structured career pathways, including for young people and those with lower formal qualifications. If firms adopt AI primarily to reduce labour costs, productivity gains may concentrate in large companies with the capital to invest in advanced systems, potentially widening wage and opportunity gaps. AI-driven recruitment tools may unintentionally disadvantage candidates with non-standard communication styles, speech impairments, or atypical career trajectories. For instance, neurodivergent candidates may be unfairly screened out by algorithms that evaluate “enthusiasm” based on tone of voice, eye contact, or linear communication patterns. Without safeguards, such AI applications could reinforce exclusion rather than reduce it.
Yet generative AI also has strong equalising potential. As a tool of augmentation, it can expand the capabilities of workers who previously faced structural barriers. A small retail entrepreneur can use AI to generate marketing content, analyse sales data, and manage inventory without hiring external specialists, lowering barriers to business growth. A freelance designer can rapidly prototype ideas and compete with larger agencies. In these cases, AI functions as a “capability multiplier,” distributing access to expertise that was once concentrated among highly educated professionals or large firms.
The inclusion potential is particularly significant for people with disabilities and special educational needs. Generative AI can operate as an assistive layer that reduces workplace barriers. Individuals with dyslexia can use AI-supported drafting and proofreading tools to strengthen written communication. Workers with speech impairments can rely on text-to-speech or AI-enhanced communication systems in remote service roles. Neurodivergent employees may use AI to structure responses, summarise meetings, or prepare for interviews. Remote and hybrid work, supported by AI automation of repetitive tasks, can also benefit individuals with mobility limitations or chronic health conditions. In these contexts, AI does not replace workers; it reduces the impact of impairment on performance, enabling skills and insight to take centre stage.
However, this inclusive scenario is not automatic. AI could simultaneously eliminate structured, predictable roles that have traditionally offered accessible employment pathways. If entry-level tasks are automated without parallel investment in reskilling, some groups may face shrinking opportunities. Moreover, access to high-quality AI tools and digital training is uneven. Without affordable infrastructure, strong digital skills development, and employer awareness, augmentation benefits may accrue primarily to already advantaged workers.
For generative AI to reduce inequality in Europe, several conditions are essential:
First, broad and affordable access to AI tools must be ensured across sectors and regions.
Second, vocational education and training (VET) and lifelong learning systems must integrate digital and AI literacy, ensuring that learners at all levels acquire the skills to use, manage, and critically evaluate AI technologies. This includes not only technical competencies, such as generating, interpreting, and applying AI outputs, but also ethical awareness, data literacy, and the ability to adapt workflows in AI-augmented environments. Crucially, these programs must provide targeted support for vulnerable groups, including people with disabilities, neurodivergent learners, individuals with limited formal education, and those from disadvantaged socio-economic backgrounds. Tailored interventions, such as accessible learning materials, assistive technologies, flexible pacing, mentorship, and personalized guidance, can help these learners benefit from AI rather than be excluded by it. By embedding digital and AI literacy into inclusive VET and lifelong learning pathways, Europe can ensure that technological progress translates into broader labour market participation, equity, and resilience, rather than reinforcing existing inequalities.
Third, inclusive design principles and regulation are needed to prevent algorithmic bias in recruitment and performance management. Fourth, AI should be recognised as a legitimate assistive technology and reasonable workplace accommodation. Finally, social dialogue and policy coordination must ensure that productivity gains translate into quality jobs rather than labour market polarisation.
Generative AI is neither inherently equalising nor inherently divisive. Its impact will be shaped by policy choices, institutional capacity, and the inclusiveness of Europe’s skills ecosystems. If aligned with inclusive VET strategies and active labour market policies, AI can help level the playing field. If left solely to market forces, it risks reinforcing existing divides. The challenge for European stakeholders is therefore not whether AI will transform work, but how to steer that transformation towards greater inclusion.
If you are leading research in this field, consider submitting to Cedefop’s call for papers on harnessing artificial intelligence and digital technologies for inclusive and resilient vocational education and training (VET). This is a unique opportunity to contribute to a Cedefop publication, participate in a high-level event in 2027, and ensure that your research reaches key stakeholders and helps shape European policy-making and the development of inclusive VET systems that promote skills, equity, and accessible employment for all.
