AI to Support Neurodivergent Learners in Vocational Education and Training

1. Introduction

This report discusses how AI (Artificial Intelligence) and other advanced technologies can support neurodivergent learners in Vocational Education and Training (VET) settings, including work-based learning and apprenticeships, and the school-to-work transition. It explores the use of these technologies by learners with Autism Spectrum Disorder (ASD), learners with Attention-Deficit/Hyperactivity Disorder (ADHD), and learners with learning disabilities, such as dyslexia, dyscalculia and dysgraphia, as well as use by teachers and trainers in catering for these learners and promoting neuroinclusive learning.

The current challenges for neurodivergent learners in VET and the transition to work

There are clear and persistent gaps in educational and employment outcomes associated with neurodivergence

Young neurodivergent people (see Box 1.1) can face a struggle to complete compulsory education and to transition to further and higher education and training and eventually the labour market. Persistent gaps in educational and employment outcomes signal that neurodiverse talents are undervalued and underused, imposing costs to society in the form of lost productivity, higher welfare spending and worse health outcomes (Mezzanotte and Calvel, 2023[1]). More inclusive education, both general and vocational, and workplaces could help bridge these gaps.

Box 1.1. Who are the neurodivergent learners discussed in this report?

Neurodiversity describes natural variations in how people think, learn, perceive the world, interact, and process information (EARN[2]). The term emerged within advocacy in the late 1990s with the purpose of shifting discourse beyond purely medical terms (Singer[3]). As a result, there is no single, standardised definition of the term. While the term “neurodivergent” can refer to a wide range of people, this study focusses primarily on learners with Autism Spectrum Disorder (ASD); learners with Attention-Deficit/Hyperactivity Disorder (ADHD); and learners with learning disabilities, such as dyslexia, dyscalculia and dysgraphia. These neurodivergences are relevant for this study because they are often diagnosed during an individual’s school years. Additionally, the focus on a narrower set of neurodivergences helps narrow the scope of technologies under consideration so that their benefits and risks can be better understood. The benefits and risks are not necessarily unique to learners with these neurodivergences; many of them also apply to learners with other neurodivergences and/or disabilities and to neurotypical learners.

Neurodivergence is a risk factor for early school leaving. Box 1.2 explains how ASD, ADHD, dyslexia, dyscalculia and dysgraphia can affect learning, while the next section outlines some challenges faced by neurodivergent learners in VET. Compared to their peers without ADHD, learners with ADHD have (Mezzanotte, 2020[4]): school attendance challenges and higher risk of dropping out of high school (Barbaresi, 2007[5]); higher risk of being expelled or suspended (LeFever, 2001[6]); and, lower likelihood of completing post-secondary education (DuPaul, 2009[7]). Students with learning disabilities are twice as likely to leave school before completing their education compared to their peers (Kaur, 2017[8]). Early school leaving adds to the challenges young people with disability face in the transition to the labour market (OECD, 2022[9]), and puts them at a disadvantage for the many key life outcomes linked to education, such as physical and mental health, social mobility, criminality, and social cohesion and political participation (Mezzanotte, 2022[10]).

Box 1.2. How neurodivergences can affect learning

Autism Spectrum Disorder (ASD) comprises a diverse group of conditions related to brain development, characterised by some degree of difficulty with social interaction and communication, as well as atypical patterns of activities and behaviours (World Health Organization, 2025[11]). The needs and abilities of individuals with ASD vary and can evolve over time. Depending on where the conditions lie along the spectrum, some people can live independently while others deal with severe impairments and require care and support their whole life. The World Health Organization estimates that around 1 in 127 people has ASD. Impairments in communication, social interactions and cognition can significantly disrupt daily functioning, developmental progress and educational and social outcomes for these children (World Health Organization, 2013[12]).

Attention-Deficit/Hyperactivity Disorder (ADHD) is a neurodevelopmental disorder characterised by a persistent pattern of inattention and/or hyperactivity-impulsivity lasting at least six months, with onset in early to mid-childhood, and a severity that exceeds typical age‑related variations, significantly impairing academic, occupational or social functioning (Mezzanotte, 2020[4]; World Health Organisation, 2025[13]). Post-2000 estimates of prevalence of ADHD diagnosis in children range from 3.2% in Sweden to 10.5% in the United States (Martin et al., 2025[14]) ADHD impacts students’ ability to function effectively in school, which can be partly explained by factors such as challenges in executive functions, like working memory, planning, organising and shifting between tasks (Wiener and Daniels, 2015[15]).

Individuals with dyslexia experience difficulties in learning academic skills related to reading and writing, such as word reading accuracy, reading fluency and reading comprehension (World Health Organisation, 2025[13]). It is estimated that around 15% to 20% of the world population have some of the symptoms of dyslexia (International Dyslexia Association, 2020[16]). Dyslexia can negatively impact individuals’ educational and professional outcomes, as well as their emotional well-being, self-esteem, and social relationships and behaviours (Livingston, Siegel and Ribary, 2018[17]).

Individuals with dysgraphia experience difficulties in learning academic skills related to writing, such as spelling accuracy, grammar and punctuation accuracy and the organisation and coherence of ideas in writing (World Health Organisation, 2025[13]). It is unclear what proportion of the population experiences dysgraphia due to variability in diagnostic criteria, methodologies and awareness (Biotteau, 2019[18]). Dysgraphia can also create barriers for the acquisition of other skills and performance of core activities (e.g. note‑taking, self-expression), as well as negatively impact students’ self-esteem and in turn, their academic success (Danna, Puyjarinet and Jolly, 2023[19]; Asselborn, Chapatte and Dillenbourg, 2020[20]).

Individuals with dyscalculia experience difficulties in learning academic skills related to mathematics or arithmetic, such as number sense, memorisation of number facts, and accurate mathematical reasoning (World Health Organisation, 2025[13]). Dyscalculia is estimated to impact 5 to 7% of the population, frequently coexists with dyslexia and can be associated with long-term academic difficulties (Kaplan and Meylani, 2025[21]).

Source: Extracts from the OECD working paper (Linsenmayer, 2025[22]), “Leveraging artificial intelligence to support students with special education needs”, https://doi.org/10.1787/1e3dffa9-en.

Indeed, there is an employment gap for neurodivergent adults. One survey in the United States suggests that adults with ADHD are 25 percentage points (p.p.) (34% vs. 59%) less likely to be employed full-time than adults without ADHD, while a longitudinal study also in the United States associates childhood ADHD with an employment gap of 10 to 14 p.p. in adulthood, an earnings reduction of approximately 33%, and greater likelihood (by 15 p.p.) of receiving social assistance (Mezzanotte, 2020[4]; Biederman and Faraone, 2006[23]; Fletcher, 2013[24]). Some studies from the United Kingdom and the United States suggest that between 50 and 75% of adults with ASD are unemployed (Jacob et al., 2015[25]).

Comparison across disability type or neurotype is challenging given high rates of co‑occurrence and the large heterogeneity even among people with the same disability or neurodivergence. However, 2022/23 official statistics for people with disability in the United Kingdom (Department for Work and Pensions[26]) suggest that autism and learning difficulties are associated with worse employment outcomes on average than many other health conditions. The employment rate is 31% among people whose main health condition is autism and the same rate for people whose main health condition is severe or specific learning difficulties. These employment rates are the lowest of all the health conditions included in the data, including for example, difficulty hearing (75%), musculoskeletal conditions (58%) and mental health conditions (44%). In the same direction, recent research found that neurodivergent people in the United Kingdom are about twice as likely to be in precarious work and more than ten times as likely to be in temporary employment as neurotypical peers (Branicki et al., 2024[27]).

While the educational attainment of people with disability has improved considerably in the past 15 years, so has the educational attainment of people without disability, with the result that people with disability remain at the same relative disadvantage. This is a main finding of the 2022 OECD report on “Disability, Work and Inclusion: Mainstreaming in All Policies and Practices” (OECD[9]), which tracks how disability education and employment gaps have evolved over the previous 15 years, although the statistics presented within are not disaggregated by neurodivergence or type of disability. Young people with disability are more likely to leave school early (i.e. without attaining any more than a lower secondary education). In the EU27, 22.5% of persons with disabilities aged 18 to 24 left school early in 2022, compared to 8.4% of young persons without disabilities (Eurostat[28]). 28.5% of young persons with disabilities (aged 15 to 29) in the EU27 were neither in employment nor in education and training (NEET) in 2022 (Eurostat[29]), compared to 10.8% for those without disabilities. Similar statistics are seen across the OECD (OECD, 2022[9]), although post-COVID data are not available for all countries.

The persistent disability gap in educational outcomes is mirrored in employment outcomes, with the employment rate of people with disability remaining much lower than that of people without disability. The average disability employment gap across EU member states was 22 p.p. in 2023 while people with disability were approximately twice as likely to be unemployed than people without disability in 2022 (15.1% vs. 7.6%) (Grammenos, 2024[30]). Gaps of similar magnitude are seen across the OECD. In 2022, the OECD (2022[9]) noted that these gaps in employment outcomes had remained stubbornly stable in most countries since 2008. However, once employed, the likelihood of a job-to-job change is relatively similar for people with disability and people without disability. This indicates the potential for people with disability and/or with neurodivergences to thrive in the labour market once they overcome initial barriers to entry.

Vocational Education and Training (VET) is an important setting for neuroinclusive learning

VET plays an important role in providing inclusive opportunities for all learners, including those who are neurodivergent, while maintaining its broader mission as a pathway of excellence and innovation (see Box 1.3 for a definition of VET). VET serves a diverse mix of learners with varied skills and work experience, with programmes offered at different levels reflecting the complexity of knowledge and skills involved (OECD, 2023[31]). VET can be an appealing option for some neurodivergent students, especially those who prefer practical learning over academic studies, due to the focus on practical skills tailored to specific industries and job roles. Furthermore, modern VET increasingly covers advanced cognitive, digital, and social competences, accommodating learners with many different interests. Work-based learning and the focus on employability skills can support learners’ transitions to the labour market and help close the employment gaps associated with neurodivergence. Vocational programmes can also provide opportunities to develop transferable competences, including literacy and numeracy skills, preparing learners for more advanced education and making them adaptable in a changing labour market.

Where countries see high participation of neurodivergent learners in VET programmes, this can be for the reasons above but can also reflect broader systemic factors, such as guidance and selection practices,1 or the limited capacity of general education to accommodate diverse learning needs. In the EU, many young people with disability choose or are oriented towards VET (Waddington, 2018[32]).2 In some OECD countries, VET accommodates learners with weaker literacy and numeracy skills than those in general education (OECD, 2023[31]). The OECD has highlighted the importance of inclusion in all areas of the education system3 – general and vocational, but also at primary and lower-secondary level – while also acknowledging the key role that VET can play in preventing early school leaving for young people with disabilities and/or with neurodivergences (2022[9]; 2023[33]).

However, the high participation of neurodivergent learners in VET programmes in some countries should not be understood as intrinsic to VET itself. Modern VET systems are increasingly designed as pathways of excellence, combining advanced technical, digital and transversal competences across a wide range of sectors, including high-demand and technology-intensive fields. Framing VET primarily as an option for learners who struggle in academic settings risks overlooking its evolving role as a key pillar of innovation, inclusion and high performance in education and work.

Regardless of what specific factors lead neurodivergent learners to VET, it remains an important setting for neuroinclusive education, playing a key role in keeping neurodivergent learners in education and preparing them for further and higher education and eventually the labour market.

Box 1.3. What is Vocational Education and Training (VET)?

VET encompasses a broad range of learning pathways that equip individuals with practical and job-specific skills. In line with the definition used in the European Union Council Recommendation on Vocational Education and Training (VET) for sustainable competitiveness, social fairness and resilience (24 November 2020), VET refers to education and training which aims to equip young people and adults with knowledge, skills and competences required in particular occupations or, more broadly, in the labour market. Usually VET programmes have work-based components, whereby students receive training in companies. VET programmes may also aim to develop transferable competences, including literacy and numeracy skills. Successful completion of VET programmes often leads to labour market-relevant vocational qualifications (e.g. certificate, diploma or title) that, in many systems, are acknowledged as occupationally oriented by the relevant national authorities and demonstrating that the individual has acquired competences to specific standards. VET programmes may also provide an entry to higher-level education and training.

VET plays a crucial role in OECD education systems, as it provides an alternative to general academic education, enhances employability, facilitates the transition from school to work and to further education and training, and meets labour market demand for skilled workers (OECD, 2023[31]).

Although VET structures vary significantly across OECD countries, upper secondary education (ISCED 3) remains the most common level for VET programmes. On average, VET programmes account for 42% of upper secondary students in OECD countries, reaching 70% in some countries such as the Czech Republic (Czechia) and Slovenia​ (OECD, 2023[31]). In some OECD countries, there is no distinct upper secondary VET pathway, with vocational learning instead integrated into general education or offered mainly at post-secondary level. Many countries offer VET at post-secondary non-tertiary (ISCED 4) and short-cycle tertiary (ISCED 5) levels. Some, such as Germany and Switzerland, have also developed higher-level VET pathways, where vocational qualifications are equivalent to bachelor’s and master’s degrees​ (ISCED 6 and above).

Work-based learning (WBL) is an essential element of VET and benefits both students and employers. Workplace environments provide access to up-to-date equipment and expose students to industry-relevant techniques and technologies, which may not always be available in school settings​. Countries with strong WBL components in their VET systems see improved labour market outcomes for graduates, including shorter job search durations, lower unemployment risks, and higher wages (OECD, 2023[31]; European Commission, 2025[34])​. Moreover, WBL enhances the alignment between vocational education and labour market needs, as employers play a direct role in shaping training opportunities.

Apprenticeships represent a structured form of WBL, combining classroom-based instruction with substantial on-the‑job training typically representing at least 50% of the programme duration. Students can alternate between school-based learning and paid employment in a company (dual model) or receive it in blocks. For example, in Norway, students spend the first two years in school and the last two with the company. Countries with well-established apprenticeship models, such as Germany and Switzerland, demonstrate how these programmes can effectively facilitate the transition from education to employment.

Yet neurodivergent learners can face challenges in VET

Neurodivergent learners can face challenges in VET, just as in general education. Health issues, including learning difficulties, have been identified as a factor in leaving VET early (Cedefop, 2016[35]). Other studies have highlighted barriers that people with disability face in accessing and completing VET programmes, such as the lack of tailored learning environments, insufficient support during the transition from education to employment (Eurofound, 2021[36]) and insufficient provision of ICT equipment (Waddington, 2018[32]) – many of which would also apply to neurodivergent learners.

Stakeholders interviewed for the study made the point that VET and education more generally are not built for neurodiversity and that no single method of instruction works for everyone. Veronika Kaska (Deputy Director, Ashtangu Vocational Rehabilitation Centre) and Colm McNamee (Cuimsiú Employability Mentor) said that education systems are designed for the “average student”, a profile which does not match any real student, not least those with neurodivergences. Teaching materials are not typically tailored according to diverse needs, according to Elisabetta Bertola (AAC Specialist Co‑ordinator, Irisbond). Despite VET being more oriented towards practical and job-specific skills, many of the challenges for neurodivergent learners in VET are the same as in other educational environments, according to Alisdair Gurling, a researcher at Wonderlab, Monash University. For instance, even trainee electricians may need to study large safety manuals.

Teachers (including trainers) in VET face a substantial challenge in trying to accommodate all needs. This challenge is likely even greater in VET than in general education, due to the diversity of learners’ skills and work experience. Nwanneka Udeka, a speech-language pathologist, highlighted that many teachers are not trained to understand diverse needs and feel that they do not have the time needed to accommodate all needs in the class. According to Thomas Köhler, who delivers VET teacher training in Germany, it is very difficult for any VET teacher to detect and cater for all needs in a typical class of 20 students. In the most recent OECD Teaching and Learning International Survey (TALIS) of (non-VET and VET) teachers at ISCED 2 and ISCED 3 levels, approximately a third of teachers report that modifying lessons for students with special education needs is a source of stress (OECD, 2024[37]). In the VET context, these challenges are often compounded by the fact that the responsibility for learner support is split between education and training providers and employers.

Like other learning environments, VET learning environments may lack the psychological safety that enables learners to explore and share ideas, especially neurodivergent learners who may experience behavioural or mental health issues (i.e. beyond the learning difficulties identified in Box 1.2). Veronika Kaska explained that neurodivergences are often intertwined with behavioural or mental health issues such as anxiety and social withdrawal, previous trauma, underdeveloped socio‑emotional skills, difficulty expressing themselves and needing support in mastering basic everyday tasks. Other interviewees pointed out that neurodivergent learners may enter VET with low confidence and internalised stigma due to prior experiences in general education, bullying or exclusion.

Neurodivergent learners can face challenges in the transition to the labour market

Interviewees described a labour market wherein the full potential of neurodiverse talents is not realised, as evidenced by the employment gaps discussed previously in this chapter. Neil Milliken (Vice President, Global Head of Accessibility & Digital Inclusion at Atos) made the point that neurodivergent people can be extremely high performing employees given the right environment and the right support. Illustrating this point, Hiren Shukla (Global and Americas Neuro-Diverse Center of Excellence Leader, EY) cited an EY survey which suggests that in inclusive environments, neurodivergent professionals’ proficiency in key skills is much higher: 31% increase in proficiency for cybersecurity skills, 20% for AI and big data skills, and 10% in strengths such as resilience, flexibility and agility (EY, 2025[38]).

Misconceptions and a lack of awareness around neurodiversity can lead employers to focus on perceived limitations rather than strengths and can place neurodivergent VET learners at a disadvantage in securing work-based learning placements. Some employers approach neurodiversity and disability with trepidation, according to Nicole Lonican, whose organisation, FIT, runs an employee network aimed at capacity building to help employers support workers with disability and neurodivergences and to provide workplaces that are accessible for everybody. She noted how employers are sometimes reluctant to engage with the topic due to a lack of awareness of how to include neurodivergent talents or a fear of offending neurodivergent employees. Francesc Sistach (CEO, Specialisterne Global) described how these fears and lack of awareness can lead companies to ultimately view neurodiversity as a cost rather than an opportunity.

Issues around diagnosis and disclosure prevent neurodivergent employees from accessing the supports they need, which could include in-work grants as well as access to assistive technologies, including AI and other advanced technologies. In Ireland, Nicole Lonican (Cuimsiú Programme Co‑ordinator, FIT) explained that long public waiting lists and the cost of private evaluations make diagnosis a privileged process, so requiring formal proof can exclude those most in need. She described how lower confidence, internalised stigma and lower self-advocacy among neurodivergent individuals are not only barriers to entry to the job market but also prevent employees from being open about their needs and seeking support. Many interviewees highlighted that stigma around neurodivergence discourages disclosure among job candidates and employees.

Many of the same issues around diagnosis and disclosure also affect VET. Interviewees described a feedback loop between VET and the labour market. Delayed diagnosis and insufficient support in VET can lead to challenging transitions to work, according to Francesc Sistach. At the same time, efforts to align VET with labour market needs mean that when employers do not prioritise inclusion, the VET system does not prioritise it either, according to Pierre Dillenbourg, Professor at the Computer-Human Interaction Lab for Learning & Instruction at the Swiss Federal Institute of Technology (EPFL).

Methodology

This report summarises the key findings of over 50 stakeholder interviews as well as a workshop bringing together additional stakeholders engaged in VET and disability policy. Stakeholder interviews lasted up to an hour and were conducted virtually. Interviewers followed a semi-structured approach, using a topic guide with pre‑determined open questions (see Annex B) but with the flexibility to select and order questions according to interviewees’ profiles and expertise, and to ask follow-up questions to elicit more in-depth insights. A targeted (but not representative) approach was used to recruit interviewees with experience and knowledge of the use of AI and other advanced technologies to support neurodiversity in VET who could deliver rich and diverse insights on this topic. Potential interviewees were identified through a literature review, professional contacts and desk research (identifying participants of relevant conferences, members of research or policy groups) and by other interviewees’ referrals. Interviewees included: developers of relevant technologies, VET teachers and trainers, vocational rehabilitation practitioners, academics, employers, disability rights advocates and policymakers. A list of the participants who agreed to be named is provided in Annex A. Interviewees naturally included some individuals who described themselves as neurodivergent and were able to contribute valuable insights based on their own experiences in education and at work. Due to challenges with outreach, no interviews were secured with current VET learners.

As a synthesis of information collected through interviews, this report has the following limitations. While efforts were made to seek diverse profiles and viewpoints, it is likely that the framing of the study around the use of AI and other advanced technologies to support neurodivergent learners in VET attracted participants who believe in this potential at the expense of participants who are sceptical of this potential or less engaged with this topic whose views are not as well represented in this report. Nevertheless all interviewees answered questions about both the risks associated with these technologies and the potential. Interviewees were asked about the use of these technologies in the context of VET and the transition to work, but many of the same tools can equally be applied in general education. In some cases, interviewees described technologies that they saw as having potential future application in VET but are not in use. Where possible, an effort has been made throughout the text to distinguish between potential and actual use. While individuals with the same neurodivergence can have different support needs and levels of independence (particularly with ASD, as discussed in Box 1.2), most of the technologies mentioned in this report are used by the neurodivergent VET learner themself, with the result that learners with less independence and potentially more complex needs are not as well represented in this report.

A repository of relevant tools accompanies this report and was compiled using desk research as well as information provided by interviewees. It lists 73 AI and other advanced technologies with relevance for neurodivergent learners in VET. This list is not exhaustive but it demonstrates the wide range of available tools and their features.

Notes

← 1.  An OECD report on building future‑ready VET (2023[31]) describes the potential for AI to improve selection practices within VET, using data‑oriented methods to identify individuals and employers that could benefit from VET participation.

← 2.  In some EU countries, there are at least some types of VET programme which are perceived, and sometimes intentionally designed, as being for learners with learning difficulties or other special education needs (Cedefop, 2016[35]). So-called special vocational schools and apprenticeship programmes can indicate that insufficient progress has been made towards inclusive schooling in general and vocational education (Waddington, 2018[32]). They exist despite research showing that students with disability perform better in mainstream settings than in segregated settings, through peer learning and better social skills (OECD, 2022[9]). Special vocational schools often do not offer official qualifications and thus do not help learners to meet entry requirements of other vocational training or apprenticeships and reduce the likelihood of entering employment compared to those in mainstream vocational education (Waddington, 2018[32]).

← 3.  In line with the UN Convention on the Rights of Persons with Disabilities (2006[61]) calling on signatories to provide inclusive schooling in both general tertiary education and in VET.