In 2021, GCHQ said it had four times the national average of dyslexic people in its apprenticeship programmes, and it was actively recruiting for people who can see what is out of place in a bigger picture. Notice what that example is not. It is not about autistic coders, which is the story the tech industry usually tells about itself. Neuroinclusion in tech has been held back by a flattering stereotype, and it is time HR and talent teams in software, data and cyber moved past it.
The stereotype that flatters and limits
The idea that “tech loves autistic people” has done some good. It made neurodiversity a business conversation and gave rise to programmes such as SAP’s Autism at Work, launched in 2013 with an aim of 1% of its workforce being autistic, which redesigned hiring around tasks rather than small talk. But the stereotype narrows what employers look for and where they look.
It casts neurodivergent people as a single profile, usually a quiet, male, autistic engineer who wants to be left alone with hard problems. That leaves out ADHD colleagues who do their best work in incident response and fast product cycles, dyslexic colleagues who bring big-picture and systems thinking, and dyspraxic colleagues who may be excellent architects but are poorly served by whiteboard performances. It also creates a ceiling. If neurodivergent people are seen as brilliant individual contributors, they are rarely seen as product leaders, engineering managers or CTOs.
The data suggests the sector is not as inclusive as its reputation. The BCS found that in 2023, 14% of UK IT specialists were disabled compared with 17% of the wider workforce, a gap of roughly 66,000 people. Disabled IT specialists earned on average 89% of what their non-disabled peers earned. Those figures cover disability broadly, but they hardly describe an industry that has cracked inclusion.
What a live coding test measures
Tech hiring has a distinctive ritual: the live coding test, often done on a whiteboard or shared screen while an interviewer watches and asks you to talk through your thinking. It feels rigorous. The evidence suggests it partly measures something else.
A study by researchers at North Carolina State University and Microsoft, presented in 2020, gave 48 computer science students the same problem in two conditions: a conventional observed interview, or alone in a private room. As one of the authors put it, people in the traditional interview “performed half as well” as those who solved it in private. The authors were clear the sample was small. Even so, the direction is hard to ignore, and the effect is likely to be sharper for candidates who are anxious under observation, who process verbally more slowly than they think, or who find it hard to narrate and code at the same time.
None of this means dropping technical assessment. It means choosing formats that measure the job. Most engineers do not write code while a stranger watches them in silence. They read documentation, look things up, think, test and ask questions in writing. A take-home task of reasonable length, a choice between a live or async format, sharing the problem type in advance, or a code review exercise in place of code from scratch will all tell you more about how someone will perform on a Tuesday afternoon.
AI is rewriting the junior roles
Many neurodiversity hiring programmes in tech have used junior roles such as software testing, quality assurance and data validation as the entry point. Those are exactly the tasks AI tools are absorbing. The Institute of Student Employers reported a 46% fall in tech-sector graduate jobs over the year to autumn 2025, with a further drop projected, as employers lean on AI for routine coding and analysis and hire experienced people instead of training newcomers.
If your neurodiversity strategy depends on a pipeline of entry-level testers, it needs rethinking. The upside is that the work left for people is increasingly about judgement, spotting what the model got wrong, deep domain knowledge and asking the awkward question. Those are areas where many neurodivergent people are strong. AI tools can also remove friction for dyslexic and ADHD colleagues around writing, summarising and organising. The risk is that firms cut the junior roles without building any other on-ramp, and quietly lose a whole stream of talent.
Neuroinclusion in tech beyond the hiring pipeline
Day-to-day tech culture has its own pressure points. Agile ceremonies stack up into back-to-back meetings. Daily stand-ups reward people who speak fluently on the spot. Slack and Teams generate a constant stream of notifications that is draining for anyone, and more so for people whose brains filter input differently. On-call rotas disrupt sleep and routine. Open-plan offices with background chatter are tough for people with sensory sensitivities. Promotion frameworks often lean on “influence” and “visibility”, which can quietly penalise people who do excellent work but do not self-promote.
Cyber security deserves a specific mention. The sector talks a lot about neurodivergent people’s pattern recognition and persistence, and GCHQ has made that case publicly. But security roles often bring vetting interviews, high-pressure incident work and rigid procedures. Inclusion there means adjusting how vetting and incident processes are run, as well as celebrating the talent.
Practical moves for tech People teams
Start by auditing your interview loop. Map each stage, ask what it measures, and replace anything that tests performance under observation with something closer to the real work. Offer every candidate the same adjustments by default, such as questions in advance, a written option and extra time, so nobody has to disclose to get a fair chance. Train interviewers to separate communication style from technical skill.
Then look at how teams operate. Set norms for asynchronous communication, protect focus time, make stand-ups optional to speak at in favour of written updates, and agree response-time expectations on chat. Review on-call rotas for flexibility. Check your promotion criteria for hidden bias towards extroversion, and track whether neurodivergent employees progress into leadership at the same rate as their peers.
Finally, refresh any specialist hiring programme for the AI era. Ask which entry roles will exist in two years, and build routes into those. This applies to a 30-person start-up as much as to a large platform company, and smaller firms often have more freedom to redesign quickly.
Looking for technical talent your current interview loop might be missing? Enna’s jobs board has more than 5,000 registered neurodivergent candidates who chose a board built for them, so they are actively looking for employers like you. Post a role for free and see who applies.
