“Can you take a quick look at this?” In a product team’s messaging channel, the request seems harmless. Yet everything is left to guesswork: should the recipient fix the bug, provide an analysis or simply acknowledge receipt? Before noon or sometime this week? For some employees, this ambiguity becomes a constant extra workload. Making workplaces more welcoming to neurodiversity often starts here: fewer unspoken assumptions, clearer guidance, without asking everyone to share their medical history.
Moving beyond the impulse to label
Neurodiversity refers to the diversity of human neurological functioning. In the workplace, discussions focus in particular on autism, attention deficit hyperactivity disorder, specific learning differences and dyspraxia. These conditions are neither interchangeable nor reducible to a list of traits. Two people with the same diagnosis may have very different, even opposing, needs.
The tech industry did not wait until 2026 to take an interest. SAP launched its Autism at Work program in 2013; Microsoft introduced an autism-focused hiring initiative in 2015. These efforts have helped raise awareness of the issue. They do not, however, establish that technology companies have become inclusive, or that autistic people have a natural aptitude for coding.
The narrative of the “atypical brain, exceptional talent” remains appealing to recruiters. Yet it pigeonholes as much as it celebrates: should someone have to be a debugging genius to deserve an accessible environment? A serious approach also considers fatigue, interruptions, unspoken social expectations and everyday working conditions. Inclusion should not depend on a promise of extraordinary performance.
The hidden cost of unwritten rules
Tech teams like to describe themselves as direct and rational. Yet their daily work relies on countless unspoken conventions: knowing when to interrupt a video call, understanding that “let’s discuss” means no, distinguishing a suggestion from an instruction. Remote work and the proliferation of communication channels have sometimes shifted these ambiguities rather than eliminated them.
Imagine a developer invited to a meeting titled “Architecture check-in.” Once there, she discovers that an immediate decision on her proposal is expected. A colleague improvises easily; she would have preferred to review the objections beforehand. This mismatch does not establish any diagnosis. Above all, it reveals inadequate preparation, which turns verbal fluency into an implicit measure of technical competence.
The right question, then, is not: “Who on this team is neurodivergent?” Instead, it becomes: “What information do people currently have to guess in order to do their jobs well?” This review of unspoken assumptions also benefits new hires, people working in a second language and colleagues experiencing temporary exhaustion.
Three adjustments that change collaboration
Before the meeting, set out the ground rules
A useful agenda specifies the objective, topics, required documents and expected outcome. “Choose between two storage options” provides better preparation than “Technical sync.” Stating whether a meeting is intended to inform, explore or decide prevents people from discovering too late that they were expected to defend a position.
Whenever possible, documents should arrive early enough to be reviewed during working hours. Participants can submit written input before the meeting or add to their response afterward, if the decision allows it. This rules out neither spontaneity nor debate. It simply avoids reserving influence for those who think aloud fastest.
In instructions, make the outcome observable
“Improve the page” leaves plenty of room for interpretation. An actionable request describes the deliverable, priority, deadline and acceptance criteria. It also distinguishes non-negotiable constraints from choices left to the individual. Precision is not micromanagement when it clarifies the goal without dictating every step.
- Deliverable: a proposed fix, not an immediate deployment.
- Scope: the mobile sign-up flow.
- Deadline: Thursday afternoon, with any blockers flagged in advance.
- Approval: a review by the technical lead before release to production.
Asking “What’s missing before you can get started?” opens up the discussion more than a simple “Is that clear?” The person receiving the request can summarize its key points; the person making it remains responsible for resolving ambiguities. Clarification must not become a comprehension test.
For feedback, replace surprises with structure
A message saying “Can we talk?” sent without context can trigger hours of worry. Specifying the subject, urgency and expected preparation reduces that uncertainty. Regular check-ins also make it possible to address difficulties before they surface, piled up, during the annual review.
Useful feedback describes a fact, its effect and the adjustment expected: “The documentation doesn’t mention the migration; add that step before approval.” This is more actionable than “You lack rigor.” Predictable does not mean watered down. Criticism can be candid without targeting someone’s personality or using an assumed diagnosis as an explanation.
Offer options without requiring disclosure
A manager can ask the whole team: “Would you prefer to receive the topics in advance?”, “Which channel works for urgent matters?” or “Do you need time to think before responding?” These questions concern work. They do not invite colleagues to investigate someone’s health.
In France, an occupational health physician can help assess needs and recommend workplace adjustments. Medical confidentiality protects health information; a manager does not need to know a diagnosis to implement recommendations. Some support schemes require an individual application, but many communication improvements can be offered to everyone without supporting documentation.
Care is needed, however, not to replace one rigid standard with another. Putting everything in writing can help some people while overloading those who have difficulty reading. A canceled meeting can turn into an endless written discussion. The aim is to offer compatible formats: a short summary, a verbal explanation, reference material and a decision recorded in the same place.
Measure work, not brains
There is no need to compile a database of presumed profiles to track progress. A team can monitor requests for clarification, tasks redone because instructions were misunderstood, meeting length or how easy it is to flag a blocker. These indicators require collective interpretation: more questions may reflect greater trust, not a deterioration.
AI tools already capable of summarizing conversations can support this approach, provided confidentiality is protected, participants are informed and outputs are checked. But an inaccurate summary can also cement a misunderstanding. These tools should never be used to infer a condition from someone’s writing style or behavior in meetings.
What now? Looking ahead to September 2026, the most useful prospect is not management that diagnoses, but management that communicates more explicitly. A team can test an agenda template, structured instructions and feedback announced in advance for a month, then make adjustments together. If these practices spread, their success will be seen less in the labels people display than in a tangible possibility: working together without having to explain how one’s brain works in order to be understood.
