When Ernst & Young Canada launched its first Neurodiversity Centre of Excellence in Toronto in 2020, the goal was straightforward: tap into an overlooked talent pool to address persistent skills shortages. Five years later, the company had hired over 2,500 neurodiverse professionals worldwide, with retention rates reaching 92 per cent. “The level of creativity, the innovation, the productivity that they are able to deliver is incredible,” one program leader told CBC News.
EY is not alone. SAP, Microsoft, JPMorgan Chase, and Hewlett Packard Enterprise (HPE) have all launched neurodiversity hiring initiatives with remarkable results. HPE’s neurodiverse testing teams achieved 30 per cent higher productivity than traditional teams. JPMorgan reported gains of 90 per cent to 140 per cent in certain technology roles. SAP’s Autism at Work program maintains retention rates exceeding 90 per cent.[1]
These programs demonstrate that neurodivergent employees, including those with autism, attention deficit hyperactivity disorder (ADHD), dyslexia, and related conditions, can deliver exceptional value when organizations design processes that recognize diverse cognitive styles. Yet despite a decade of documented success, these initiatives remain exceptions rather than the norm. The main reason lies not in organizational resistance to inclusion, but in a technological contradiction: while a handful of companies learned to hire neurodivergent talent through redesigned processes, most organizations simultaneously deployed AI systems that screened them out.
The Scale of Exclusion
The employment gap facing neurodivergent individuals persists across cultural contexts, from Western economies to East Asia, even in jurisdictions that have embraced equity, diversity, and inclusion as policy priorities. In Canada, Statistics Canada reports that only 33 per cent of adults with autism spectrum disorder are employed, compared to 79 per cent of adults without disabilities.[2] The employment rate for persons with disabilities overall dropped to 46.4 per cent in 2024, with youth aged 15 to 24 experiencing a particularly sharp decline of 6.9 percentage points.[3] The Azrieli Foundation estimates that for Canadians with ADHD, dyslexia, autism, or other neurodivergent conditions, the employment rate falls to just 26 per cent.[4] According to the Government of Canada’s 2022 Canadian Survey on Disability, over one million persons with disabilities aged 15 to 64 were out of the workforce, even though they could work if provided the right supports.[5]
The pattern holds in the United States. The American Enterprise Institute reports that unemployment rates for neurodivergent adults run as high as 30–40 per cent, roughly eight times the rate for people without disabilities.[6] Some estimates place unemployment among adults with autism spectrum disorder at 85 per cent, even among those with college degrees.[7] Researchers estimate that the neurodivergent Americans population totalling 67 million people represent the country’s largest untapped talent pool. One in 36 American adults has autism spectrum disorder, one in 22 has ADHD, and between 3 per cent and 7 per cent have dyslexia.[8]
In Taiwan, a jurisdiction that has made significant investments in both AI development and inclusive employment policy, Ministry of Health and Welfare data from 2025 shows 23,055 individuals registered with autism spectrum conditions. Notably, over 76 per cent (17,566 individuals) are classified as mild severity, representing a substantial population capable of workforce participation with appropriate support.[9] The distribution pattern suggests that the vast majority of autistic individuals in Taiwan could contribute meaningfully to the economy if systemic barriers were addressed.
These figures from across Western and East Asian contexts point to a troubling reality: neurodivergent individuals, who represent 15 to 20 per cent of the global population according to the World Economic Forum, face systematic exclusion from economic participation regardless of local policy commitments to inclusion.[10] The economic cost is substantial. Canadian researchers estimate annual losses ranging from C$6 billion to C$11 billion due to neurodivergent unemployment.[11] A study cited by Gitnux estimated the cost of autism alone to the U.S. economy at US$268 billion annually, primarily due to lost productivity.
When AI Becomes the Barrier
The hiring landscape has transformed dramatically over the past decade. An estimated 70 per cent of companies and 99 per cent of Fortune 500 firms now use AI tools in recruitment, according to a report in The Hill. These systems promise to eliminate human bias and improve efficiency. The evidence suggests they often achieve the opposite.
AI recruitment tools learn from historical data. If past successful hires were predominantly neurotypical, the algorithm learns to favour neurotypical patterns. Résumé-screening systems penalize non-linear career paths common among neurodivergent workers who may have experienced employment gaps or career changes. Video interview analyzers flag atypical eye contact, facial expressions, or speech patterns that are characteristic of autism. Personality assessments reward responses that cluster around neurotypical norms, systematically disadvantaging those whose cognitive styles differ from the majority.
Evidence from résumé-screening research suggests that these concerns are not hypothetical. Research from the University of Washington, presented at the 2024 AAAI/ACM Conference on AI, Ethics, and Society, examined how large language models ranked over 500 real-world résumés with names varied to reflect different racial and gender associations. The models favoured white-associated names 85 per cent of the time, favoured female-associated names only 11 per cent of the time and never favoured Black male-associated names over white male-associated names.[12] The point is not that racial bias and neurodivergent exclusion are identical, but that résumé-screening systems can convert proxy signals into unequal outcomes. That same mechanism is especially troubling for neurodivergent applicants, whose employment gaps, non-linear career histories, disability-related credentials, communication styles, or assessment responses may be read by AI systems as deficits rather than differences. Previous studies also found ChatGPT exhibits both racial and disability bias when sorting résumés.[13] As lead researcher Kyra Wilson observed, “The use of AI tools for hiring procedures is already widespread, and it’s proliferating faster than we can regulate it.”[14]
A 2024 survey found that 76 per cent of neurodivergent job seekers feel that traditional recruitment methods, including timed assessments and panel interviews, put them at a disadvantage. This challenge is compounded by the finding that 68 per cent of HR professionals acknowledge that their recruitment frameworks are not tailored to highlight the unique strengths of neurodivergent applicants.[15]
The legal landscape is beginning to respond. In 2024, a lawsuit against Workday Inc. alleged that its algorithm-based screening software discriminated against job applicants based on race, age, and disability, violating Title VII of the Civil Rights Act, the Age Discrimination in Employment Act, and the ADA Amendments Act. The plaintiff, an African-American man over forty with a disability, argued that his race was determinable because he had graduated from a historically Black college, his age was determinable from his graduation year, and his mental disabilities could be revealed through personality tests embedded in the system.[16] When the case survived a motion to dismiss in July 2024, it signalled that courts were beginning to take algorithmic discrimination seriously.[17]
New York City’s Local Law 144, effective since January 2023, became the first U.S. regulation to require annual bias audits for automated hiring tools, public disclosure of audit results, and notification to applicants about AI use.[18] The European Union’s AI Act classifies hiring AI as “high-risk,” requiring strict audits and transparency.[19] A November 2024 report from the UK Information Commissioner’s Office raised concerns that AI recruitment software was filtering candidates based on protected characteristics including gender, race, and sexual orientation, and that systems designed to monitor bias could inadvertently infer candidates’ protected characteristics using application data.[20]
These developments reflect growing recognition that AI hiring tools, often deployed as anti-bias interventions, may actually entrench and obscure discrimination. As University of North Carolina law professor Ifeoma Ajunwa, founding director of UNC’s AI Decision-Making Research Program, told NPR: “There is now a recognition of how these tools, which are usually deployed as an anti-bias intervention, might actually result in more bias – while also obfuscating it.”[21]
Beyond Screening: The Performance Management Problem
Hiring is only the entry point. The same algorithmic logic now permeates performance management, creating ongoing barriers for neurodivergent employees who successfully navigate initial screening.
Modern workplace analytics platforms continuously monitor communication patterns, collaboration metrics, and productivity rhythms. They generate engagement scores, flag employees for intervention, and recommend development actions. Integrated systems combine behavioural monitoring with generative AI to produce personalized feedback, coaching recommendations, and performance assessments. For employees whose work patterns align with neurotypical expectations, these systems function as intended, providing actionable insights that support professional development. For those with different cognitive styles, these systems can systematically misinterpret strengths as deficits.
Consider the employee with ADHD whose productivity follows a cyclical pattern: periods of intense hyperfocus followed by comprehensive output, then recovery before the next cycle. A monitoring system calibrated to expect steady, consistent engagement may flag this pattern as “inconsistent performance” or “variable engagement.” The employee delivers excellent work, but the algorithm interprets their natural rhythm as a problem requiring intervention.
Or consider the autistic employee whose preference for thorough, asynchronous communication diverges from real-time collaboration norms. They may take longer to respond to messages because they are crafting careful, comprehensive replies. They may prefer email to spontaneous video calls because written communication allows them to process information at their own pace. The system may recommend “communication coaching” or flag them for “low responsiveness” despite consistently high-quality deliverables.
“The AI may function exactly as designed. Yet the design itself embeds assumptions about optimal performance, and those assumptions typically reflect neurotypical patterns presented as universal standards.”
Research on how different neurotypes engage with technology offers an instructive parallel. A 2025 study published in Medicine examined adolescents with gaming disorder alongside ADHD or social anxiety disorder comorbidities.[22] The researchers found that adolescents with gaming disorder showed different patterns depending on whether they also had ADHD or social anxiety. Those with ADHD were more likely to be drawn to achievement within games, such as rewards, progress, and competition, and also showed higher impulsivity and more difficulty managing emotions. Those with social anxiety were more likely to use gaming as an absorbing escape from real-world stress or discomfort, a pattern associated with lower self-esteem, lower confidence in school, social, and emotional situations, and lower perceived parental support.
The clinical implications were clear: effective intervention requires understanding these as distinct profiles requiring different approaches, not as variations of a single disorder to be treated uniformly. The researchers emphasized that “the clinical approach to an adolescent with GD (gaming disorder) must be tailored according to individual characteristics.”[23]
The same principle applies to workplace AI systems. Systems designed for a single cognitive profile will inevitably disadvantage those who process information, manage attention, and engage socially in different ways. An employee with ADHD may need different performance metrics than an employee with autism, who may need different metrics than a neurotypical employee. Yet most workplace AI systems apply uniform standards derived from neurotypical baselines.
The AI may function exactly as designed. Yet the design itself embeds assumptions about optimal performance, and those assumptions typically reflect neurotypical patterns presented as universal standards.
The Business Case Remains Compelling
Despite these barriers, evidence for the organizational value of neurodiversity continues to accumulate across multiple dimensions.
A 2024 industry study found that 63 per cent of companies with neuroinclusive hiring practices saw improvements in overall employee well-being, 55 per cent observed stronger company culture, and 53 per cent reported better people management. Notably, 89 per cent of organizations that adopted neuroinclusive practices reported uplift in employee morale and engagement across their entire workforce, not just among neurodivergent employees.[24] Inclusion, it appears, benefits everyone.
Research has documented that neurodivergent employees often demonstrate specialized strengths in pattern recognition, systematic analysis, sustained attention to detail, and resistance to social conformity pressures.[25] These capabilities prove particularly valuable in roles requiring precision, analytical rigour, or independent judgment.
A 2023 study published in Autism Research found that autistic employees were more likely than neurotypical peers to identify inefficiencies and problematic practices, effectively functioning as organizational early-warning systems.[26] The researchers attributed this to reduced susceptibility to social influence and greater willingness to raise concerns regardless of interpersonal consequences. In effect, autistic employees were more likely to act as “canaries in the coal mine,” flagging problems that others might overlook or avoid mentioning.
A 2024 York University study demonstrated that autistic employees reported significantly lower levels of moral disengagement in workplace scenarios.[27] Moral disengagement refers to the cognitive mechanisms people use to rationalize unethical behaviour, such as , displacing responsibility, or dehumanizing victims. The finding suggests that autistic employees may be less susceptible to some of the rationalizations that allow workplace misconduct to persist, pointing to potential value in compliance, auditing, internal controls, and ethical oversight roles.
The cognitive diversity literature reinforces these findings at the team level. Research demonstrates that groups drawing on varied information-processing approaches, heuristics, and mental models consistently outperform more homogeneous teams on complex problems.[28] Cognitively diverse groups can even outperform teams composed of individually superior problem-solvers because high-ability individuals often converge on similar approaches, while diverse teams draw on complementary perspectives that challenge assumptions and surface blind spots.[29]
The innovation benefits are substantial. A 2018 Deloitte report found that companies with inclusive cultures were six times more likely to be innovative and agile.[30] Microsoft’s Neurodiversity Hiring Program has placed over 200 full-time employees since its inception, with the company reporting that neurodiverse teams outperform neurotypical ones on productivity and innovation metrics.[31]
Yet despite this evidence, only about 30 per cent of Fortune 500 companies have launched formal neurodiversity hiring initiatives.[32] Fewer still have examined whether their AI systems undermine those initiatives by screening out the very candidates they seek to attract, or by systematically misinterpreting neurodivergent work patterns once employees are hired.
What Successful Programs Have Learned
The neurodiversity programs that succeeded did so by fundamentally rethinking processes rather than retrofitting accommodations onto existing systems. Their experiences offer lessons for addressing AI-related barriers.
SAP, which launched its Autism at Work program in 2013, replaced traditional interviews with multi-week assessments. For years, candidates worked with Lego Mindstorms robotics kits on progressively complex projects, first individually and then in teams. This extended evaluation period allowed capabilities to emerge that would never surface in a conventional interview. As one SAP executive observed, traditional hiring processes using the same approach for everyone would systematically miss autistic candidates regardless of their qualifications.[33]
Microsoft has partnered with universities to create pathways that bypass conventional application processes entirely. Rather than screening résumés and conducting interviews, the company identifies candidates through extended engagement programs where both parties can assess fit over time.[34]
EY built dedicated Neurodiversity Centres of Excellence with specialized support structures, working with research and academic partners to develop evidence-based hiring methodologies. The program began as a talent initiative to access an untapped market but evolved into a broader transformation of how the company thinks about cognitive diversity.[35]
These approaches share several common elements. First, they extend the evaluation timeline, recognizing that neurodivergent candidates may need more time to demonstrate their capabilities than traditional processes allow. Second, they emphasize demonstration over description, creating opportunities for candidates to show what they can do rather than relying on their ability to articulate their qualifications verbally. Third, they involve neurodiversity expertise and partnering with specialists, advocacy organizations, and neurodivergent individuals themselves to design processes that work.
Fourth, and perhaps most importantly, they question the assumption that existing processes are neutral. The conventional interview, the standard personality assessment, the typical collaboration metric: these are not objective measures of capability. They are cultural artifacts that happen to favour neurotypical cognitive styles. Recognizing this opens the door to designing alternatives.
Three Questions for Leaders
The path forward does not require abandoning AI. It requires asking different questions about how AI systems are designed, deployed, and monitored. Three questions can guide this inquiry.
First, what assumptions about optimal performance are embedded in your AI systems? Most performance analytics platforms are calibrated against historical data from predominantly neurotypical workforces. The algorithms learn what “good” looks like from past examples, and if those examples reflect neurotypical patterns, the system will treat neurotypical patterns as the standard against which all employees are measured.
If your system flags “inconsistent engagement patterns,” ask whether those patterns might reflect acceptable cognitive variation rather than performance problems. If it identifies “atypical communication styles,” consider whether the baseline for “typical” reflects genuine job requirements or merely majority preferences. Audit your AI systems for neurotypical bias with the same rigour you would apply to auditing for racial or gender discrimination.
This is not merely an ethical imperative. As the Workday litigation demonstrates, algorithmic discrimination based on disability is increasingly attracting legal scrutiny. Organizations that fail to examine their AI systems for neurotypical bias may find themselves exposed to liability they have not anticipated.
Second, who participated in designing and testing your AI tools? Research consistently shows that AI systems perform better for populations represented in their development process. A 2024 survey found that only 28 per cent of HR professionals are “very confident” in identifying neurodivergent conditions, while 9 per cent are “not at all confident.”[36] If neurodivergent users and neurodiversity experts were not involved in development and testing, the system almost certainly contains blind spots.
This gap can be addressed. Engage neurodivergent employees in reviewing your AI tools. Partner with neurodiversity advocacy organizations and academic researchers who specialize in inclusive design. Create feedback mechanisms that allow neurodivergent employees to report when AI systems produce assessments that seem inconsistent with their actual performance. Treat these reports not as complaints but as valuable data about system limitations.
The same principle of inclusive participation that distinguished successful neurodiversity hiring programs applies to AI design. SAP worked with Specialisterne, a Danish organization founded specifically to help companies employ neurodivergent talent. EY collaborated with research institutions and autism advocacy groups. These partnerships brought knowledge that internal teams lacked and helped identify problems that would otherwise have gone unnoticed.
Third, what recourse exists when AI-influenced decisions prove wrong? When an algorithm contributes to a negative performance review, a rejected application, or a missed promotion, can the affected individual understand what happened and challenge the outcome? Accountability requires transparency, and transparency requires designing systems that can explain their reasoning in terms humans can understand.
This is particularly important for neurodivergent employees, who may face additional barriers in navigating traditional grievance and advocacy processes. An employee who struggles with verbal communication may find it difficult to contest an algorithmic assessment in a meeting with HR. An employee who processes information differently may need more time or different formats to understand how the system evaluated them.
Effective recourse therefore requires more than a single grievance channel. The organizations that succeeded with neurodiversity hiring created multiple pathways for employees to raise concerns and receive support. They assigned mentors, coaches, and advocates who could help neurodivergent employees navigate organizational systems. They recognized that providing equitable access to recourse mechanisms might require providing different types of support to different employees.
From Accommodation to Cognitive Pluralism
The traditional approach to disability in the workplace centres on accommodation: identifying individual employees who need adjustments and providing those adjustments so they can function within existing systems. This framework has value, but it treats neurodivergent employees as exceptions to be managed rather than as legitimate members of a cognitively diverse workforce.
The neurodiversity programs that succeeded moved beyond accommodation toward what might be called cognitive pluralism: designing systems that recognize diverse cognitive approaches as legitimate from the outset rather than treating them as deviations from a neurotypical norm.
The distinction matters. An accommodation mindset asks, “How can we help this neurodivergent employee fit into our existing system?” A cognitive pluralism mindset asks, “How can we design systems that work for employees with different cognitive styles?” The first approach retrofits adjustments after the fact. The second builds flexibility into the foundation.
Applied to AI systems, cognitive pluralism means several things. It means designing performance metrics that can recognize multiple patterns of effective work, not just the patterns that happen to characterize neurotypical employees. It means building feedback systems that can communicate in different modalities and at different levels of detail, so that employees can access information in whatever format works best for them. It means creating algorithms that treat variation as signal rather than noise, recognizing that an employee whose patterns differ from the majority might be contributing something valuable precisely because they approach work differently.
This is harder than it sounds. It requires questioning assumptions so deeply embedded in organizational practice that they feel like natural law. The idea that steady, consistent productivity is better than variable productivity seems self-evident until you consider that breakthrough insights often emerge from the kind of deep, immersive focus that naturally alternates with periods of lower intensity. The idea that rapid, real-time communication is superior to slower, asynchronous exchange seems obvious until you recognize that some of the most thoughtful contributions come from people who need time to process before they respond.
But the companies that figured out neurodiversity hiring learned that what feels natural often reflects convention rather than necessity. SAP discovered that interviews are a terrible way to assess analytical talent. Microsoft learned that conventional application processes screen out exceptional candidates. EY found that providing specialized support structures yielded retention rates far above industry norms.
The Path Forward
Indeed’s Hiring Lab reported in March 2025 that the share of U.S. job postings mentioning neurodiversity-related keywords has risen from 0.5 per cent in January 2018 to 1.3 per cent in December 2024. Excluding care-related occupations, the share tripled from 0.1 per cent to 0.3 per cent over the same period.[37] Engineering roles—including civil, electrical, and industrial engineering—are among those with the highest share of postings mentioning neurodiversity keywords.
This growth is encouraging but remains modest. The vast majority of job postings make no mention of neurodiversity, and the presence of inclusive language in a posting does not guarantee that hiring processes or workplace systems will actually accommodate cognitive diversity.
The question for organizational leaders is whether they will take deliberate action or wait for external pressure to force change. The regulatory environment is tightening, with New York City, the European Union, and other jurisdictions imposing new requirements on AI hiring tools. The legal environment is evolving, with cases like the Workday lawsuit testing the boundaries of algorithmic discrimination law. The competitive environment continues to reward organizations that can access talent their competitors overlook.
More fundamentally, the technology itself is not standing still. AI systems are becoming more sophisticated, more pervasive, and more consequential for the employee experience. Performance management platforms are incorporating generative AI capabilities that can produce personalized feedback, development recommendations, and career guidance. These systems have enormous potential to support employee growth, but they also have enormous potential to encode and amplify neurotypical bias at unprecedented scale.
The organizations that thrive in this environment will be those that treat cognitive diversity not as a compliance obligation or a charitable gesture but as a genuine source of competitive advantage. They will design AI systems that recognize different cognitive styles as legitimate. They will create pathways for neurodivergent talent to enter, contribute, and advance. They will build cultures where the question “What is the best way to work?” admits multiple valid answers.
The alternative is to allow algorithmic systems to solidify neurotypical dominance, thereby screening out neurodivergent candidates, misinterpreting neurodivergent work patterns, and gradually homogenizing the cognitive composition of the workforce. This path sacrifices the innovation benefits of cognitive diversity while exposing organizations to mounting legal and reputational risk.
The companies that pioneered neurodiversity hiring demonstrated that inclusion and excellence are not trade-offs. They found that designing for cognitive diversity improved outcomes for everyone, neurodivergent and neurotypical alike. The challenge now is to apply that insight to the AI systems that increasingly mediate how work is assigned, monitored, and evaluated.
The talent is there. The business case is proven. Will organizations design systems that unlock human potential in all its variety, or will they settle for algorithms that recognize only one way to think?
References
[1] Robert D. Austin and Gary P. Pisano, “Neurodiversity as a Competitive Advantage,” Harvard Business Review 95, no. 3 (May–June 2017): 96–103.
[2] Statistics Canada, Canadian Survey on Disability, 2017 (Ottawa: Statistics Canada, 2017).
[3] Statistics Canada, “Labour Market Characteristics of Persons with and Without Disabilities, 2024,” The Daily, May 14, 2025, https://www150.statcan.gc.ca/n1/daily-quotidien/250514/dq250514b-eng.htm.
[4] Naomi Azrieli, “Embracing Neurodiversity Will Build a More Prosperous Canada,” The Azrieli Foundation, December 2023, https://azrielifoundation.org/media/embracing-neurodiversity-will-build-a-more-prosperous-canada-but-the-onus-on-inclusivity-falls-upon-all-of-us.
[5] Employment and Social Development Canada, Employment Strategy for Canadians with Disabilities: Annual Progress Report 2024–2025 (Ottawa: ESDC, 2025), https://www.canada.ca/en/employment-social-development/programs/disability-inclusion-action-plan/reports/employment-strategy-progress-report-2024-2025.html.
[6] American Enterprise Institute, “Embracing Neurodiversity at Work: Unleashing America’s Largest Untapped Talent Pool,” April 2024, https://www.aei.org/research-products/report/embracing-neurodiversity-at-work-unleashing-americas-largest-untapped-talent-pool.
[7] National Autism Association, quoted in Creative Spirit, “22 Statistics About Neurodiversity and Employment,” May 2024.
[8] American Enterprise Institute, “Embracing Neurodiversity at Work.”
[9] Taiwan Ministry of Health and Welfare, Disability Statistics, Q1 2025 (Taipei: MOHW, 2025), https://www.mohw.gov.tw
[10] World Economic Forum, “Explainer: What Is Neurodivergence? Here’s What You Need to Know,” October 2022, https://www.weforum.org/stories/2022/10/explainer-neurodivergence-mental-health.
[11] Reboot Plus, “Bridging the Employment Gap for Neurodivergent Individuals in Canada,” March 2024, https://rebootplus.ca/blog/bridging-the-employment-gap-for-neurodivergent-individuals-in-canada.
[12] Kyra Wilson and Aylin Caliskan, “Gender, Race, and Intersectional Bias in Resume Screening via Language Model Retrieval,” Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society 7, no. 1 (2024): 1578–1590.
[13] University of Washington News, “AI Tools Show Biases in Ranking Job Applicants’ Names According to Perceived Race and Gender,” October 31, 2024, https://www.washington.edu/news/2024/10/31/ai-bias-resume-screening-race-gender
[14] Kyra Wilson, quoted in University of Washington News, “AI Tools Show Biases in Ranking Job Applicants’ Names.”
[15] MyDisabilityJobs, “Neurodiversity in the Workplace | Statistics | Update 2025,” August 2025, https://mydisabilityjobs.com/statistics/neurodiversity-in-the-workplace.
[16] Derek Mobley v. Workday Inc., No. 3:23-cv-00770 (N.D. Cal. 2024).
[17] Evan Gigante and Elizabeth Young, “Job Applicant’s Algorithmic Bias Discrimination Lawsuit Survives Motion to Dismiss,” Law and the Workplace, July 26, 2024, https://www.lawandtheworkplace.com/2024/07/job-applicants-algorithmic-bias-discrimination-lawsuit-survives-motion-to-dismiss.
[18] New York City, Local Law 144 of 2021 (effective January 2023).
[19] European Union, Artificial Intelligence Act (2024).
[20] UK Information Commissioner’s Office, Report on AI in Recruitment (November 2024).
[21] Ifeoma Ajunwa, quoted in NPR, “U.S. Warns of Discrimination in Using AI to Screen Job Candidates,” May 12, 2022, https://www.npr.org/2022/05/12/1098601458/artificial-intelligence-job-discrimination-disabilities.
[22] Sarper İçen, Yasemin Taş Torun, and Yasemen Işık, “Gaming Disorder with Attention-Deficit/Hyperactivity Disorder and Social Anxiety Disorder Comorbidities: A Cross-Sectional Analysis of Differences,” Medicine 104, no. 29 (2025).
[23] İçen, Taş Torun, and Işık, “Gaming Disorder with Attention-Deficit/Hyperactivity Disorder.”
[24] City & Guilds Foundation, Neurodiversity Index (2024), cited in MyDisabilityJobs, “Neurodiversity in the Workplace | Statistics | Update 2025.”
[25] Austin and Pisano, “Neurodiversity as a Competitive Advantage,” 96–103.
[26] Lorne M. Hartman, Mehrdad Farahani, Alexander Moore, Ateeya Manzoor, and Braxton L. Hartman, “Organizational Benefits of Neurodiversity: Preliminary Findings on Autism and the Bystander Effect,” Autism Research 16, no. 10 (2023): 1989–2001.
[27] Lorne M. Hartman and Braxton L. Hartman, “An Ethical Advantage of Autistic Employees in the Workplace,” Frontiers in Psychology 15 (2024), article 1364691.
[28] Scott E. Page, The Diversity Bonus: How Great Teams Pay Off in the Knowledge Economy (Princeton: Princeton University Press, 2017).
[29] Lu Hong and Scott E. Page, “Groups of Diverse Problem Solvers Can Outperform Groups of High-Ability Problem Solvers,” Proceedings of the National Academy of Sciences 101, no. 46 (2004): 16385–16389.
[30] Deloitte, The Diversity and Inclusion Revolution: Eight Powerful Truths (Deloitte Insights, 2018).
[31] Microsoft Unlocked, “The ‘Future’ of Work Is Neurodiverse,” March 2025, https://unlocked.microsoft.com/mentra-neuroinclusion.
[32] Universal Workforce Institute, cited in MyDisabilityJobs, “Neurodiversity in the Workplace | Statistics | Update 2025.”
[33] Austin and Pisano, “Neurodiversity as a Competitive Advantage,” 96–103.
[34] Microsoft Neurodiversity Hiring Program, described in Austin and Pisano, “Neurodiversity as a Competitive Advantage,” and Microsoft Unlocked, “The ‘Future’ of Work Is Neurodiverse.”
[35] Tammy Morris, quoted in The Globe and Mail, “Why Some Companies Are Actively Recruiting an ‘Untapped Market’ of Neurodiverse Workers,” July 9, 2024.
[36] City & Guilds Foundation, Neurodiversity Index.
[37] Indeed Hiring Lab, “March 2025 US Labor Market Update: Neurodiversity Inclusive Postings Are Rising, but Untapped Potential Remains,” March 18, 2025, https://www.hiringlab.org/2025/03/18/neurodiversity-inclusive-postings-are-rising.
