About the author: Dr Caroline Gorden is the founder of CESI Learning and a published criminologist specialising in how credibility and guilt are constructed in formal processes. She holds a PhD in Criminology, has published over 25 peer-reviewed works on social exclusion, criminal justice, and institutional fairness, and is the author of Case Studies of Famous Trials and the Construction of Guilt and Innocence, published by Bristol Policy Press in 2022. She is a Visiting Research Associate at Wrexham University and has worked as a workplace coach specialising in adults with ADHD and autism since 2023. Her work at CESI Learning applies criminological research directly to workplace investigations, grievances, and disciplinary procedures.
This article is for: HR professionals, managers, and anyone who conducts or oversees workplace investigations, grievance procedures, or disciplinary interviews. It explains why guilt can be constructed through the process itself, not just by the evidence, and what that means for autistic employees in particular. Tribunal claims involving neurodivergent employees are rising sharply, and the problem often lies not in what happened, but in how the process was handled.
I want you to think about a time when you had to sit in a formal meeting and account for yourself.
Perhaps it was a performance review that had taken an unexpected turn, or a complaint had been made, or you had been asked, without much notice, to explain something that happened weeks ago. You knew you had not done anything wrong, but the moment you sat down, you felt you were being watched, and you chose your words carefully. You may have come across as guarded or overly brief, not because you were hiding anything, but because the room itself had changed the way you were communicating.
Now consider how that experience might feel for an autistic employee, and consider that it is not a one-off for them, but a pattern that repeats across formal interactions throughout their working life. Every time high stakes and formal structure combine, the same dynamics play out, and the consequences are rarely small.
The problem is not always the evidence, it is the process
Most people assume that guilt in a formal workplace process is established through evidence: what was said, what was done, what can be verified. And of course, evidence matters. But in both criminal justice and workplace settings, my research shows something more uncomfortable, which is that guilt is also constructed. It is shaped by perception, by communication style, by how someone comes across in the room, and by whether their behaviour matches what the person asking the questions expects to see.
I have spent fourteen years researching this in academic contexts and the past three years observing it directly in workplace coaching with autistic adults. The mechanisms are consistent, and the consequences for the people involved are serious.
What credibility actually measures, and what it does not
In workplace investigations, credibility is assessed constantly, often without anyone acknowledging that it is happening. Does this person seem honest? Are they being evasive? Do they seem remorseful? Are they consistent? These are not neutral assessments. They are filtered through the expectations of the person doing the assessing, and in most workplace contexts, those expectations are neurotypical ones.
Autistic employees may communicate in ways that do not match those expectations, pausing longer before answering, speaking with a precision that sounds overly literal, showing limited facial expression, or maintaining eye contact patterns that differ from what the interviewer expects. They may also recall events non-chronologically, not because they are confused about what happened, but because autistic episodic memory does not always retrieve information in a neat timeline, and under pressure they may become more compliant than the situation seems to warrant, agreeing with suggestions rather than pushing back because of anxiety and prior experience rather than any admission of guilt.
None of these things are indicators of dishonesty, but in a formal interview setting, without adequate awareness of autistic communication, they can easily be read as evasiveness, lack of remorse, or guilt. And that is the thing about credibility: it is not a measure of truth so much as a measure of how well someone performs truth in the way the assessor has been conditioned to expect it to look.
The research: when vulnerability is mistaken for guilt
In my book, Case Studies of Famous Trials and the Construction of Guilt and Innocence, published by Bristol Policy Press in 2022 and co-authored with Christopher Birkbeck, I examine how guilt and innocence in formal processes are shaped not only by physical evidence but by sociological and psychological factors, including how someone presents, how they communicate, and how they are perceived by those in authority.
One of the cases I analyse is that of Brendan Dassey, a young man with learning differences whose confession to a serious crime was taken by investigators as proof of guilt. His behaviour during questioning was read as closed off and suspicious, and yet what those conducting the interview interpreted as evasion or suppressed guilt was, on closer examination, the response of a vulnerable young person who was overwhelmed and whose neuro-differences had never been adequately considered by those interviewing him.
The mechanisms at work in that case operate in workplace disciplinary procedures too. Research by Chandler et al. in 2018 found that autistic individuals are significantly more likely to comply with authority figures in formal settings and to agree with accusations made against them, not because they are admitting guilt, but because of learned responses to authority, high anxiety, and prior experiences of being disbelieved. These behaviours are not confessions, but they are frequently treated as such.
The case of Mr Wright: what happens when process fails
This is not a theoretical concern. It has material consequences for employees, for organisations, and for the HR professionals and managers who conduct investigations in good faith but without adequate preparation.
In the case of Mr Wright, a teacher who was awarded £850,000 in damages following a wrongful disciplinary process, the tribunal found that his communication style, which was shaped by autism, had been misinterpreted as unprofessionalism. The school had not considered whether his behaviour was a manifestation of his disability, had not reviewed what reasonable adjustments were in place, had not adapted the interview process, and had escalated through formal procedures when what was actually needed was a different kind of conversation altogether.
The result was a breach of the Equality Act 2010, significant harm to the employee, and a very costly outcome for the organisation. What that case, and many others like it, demonstrates is that the failure point is rarely malicious. Rather, it is structural, in the sense that the process itself constructs an unfair outcome even when no one involved intended that, and it is that structural dimension that makes process design so important.
What you should ask before any formal process begins
Before initiating a disciplinary procedure involving a neurodivergent employee, the first question is not “what did they do?” but rather “could this behaviour be a manifestation of disability?” If the answer might be yes, and in many cases it might be, then there are further questions that need to be worked through: what reasonable adjustments were in place before this situation arose, whether communication needs were identified and supported, whether the environment contributed to the behaviour in question, and whether the employee was given the support they needed to meet the expectations being placed on them.
These are not procedural niceties. They are the difference between a fair process and one that a tribunal will later find wanting.
If an investigation is necessary, the process must be designed for fairness
When a formal interview does need to take place, the structure of that interview matters as much as its content. Communication should be clear and in plain, accessible language, and this needs to happen in advance rather than on the day itself. Providing written notice of the topics or questions that will be covered gives the employee the opportunity to prepare, which is particularly important for autistic employees who may need additional processing time. The environment also matters: lighting, noise, seating, and the number of people in the room all affect how an autistic person is able to communicate, and these factors should be considered as reasonable adjustments rather than optional extras.
Additional time should be built into the interview itself, both for processing and between questions. Questions should be open and specific rather than leading or ambiguous, and responses should be recorded verbatim rather than paraphrased, because paraphrasing introduces interpretation at precisely the stage when accuracy is most important. The employee should be supported in choosing a companion who understands their communication style rather than simply whoever is available, and it is worth considering whether the standard interview format is the right approach at all. Structured interviewing methods such as WAFA, or Witness-Aimed First Account, are specifically designed to support clearer recall and reduce the risk of misjudgement, and they are worth knowing about.
These steps do not compromise the investigation. They are what makes it defensible.
Why I do not use the language of unconscious bias
You may have noticed that I do not frame this as an issue of unconscious bias, and there is a reason for that. Research suggests that telling people they are unconsciously biased can increase defensiveness and, in some cases, entrench the very behaviours it is intended to address, partly because it locates the problem in the individual interviewer rather than in the structure they are working within. The issue is not primarily one of individual attitudes or blind spots, even though those exist and matter; it is one of process design, and that is a more tractable and more useful place to focus.
What CESI Learning focuses on, rather than self-examination as an end in itself, is redesigning the process so that it works fairly for everyone regardless of how they communicate. Fair processes do not come about by accident, but rather through deliberate design, and that is precisely what this work is about.
Who this is for, and what to do next
If you conduct or oversee workplace investigations, grievance procedures, or disciplinary interviews, in any sector and in any role, this is directly relevant to your practice. You do not need to have made a mistake to benefit from this. As an HR professional or manager, you may well come to this course proactively, because you want to get this right before a difficult case lands on your desk, because you want your decisions to be defensible, and because you understand that the gap between good intentions and fair process is wider than most people realise.
Effective Interviewing of Autistic Employees in Investigations, Grievances, and Disciplinaries is a four-hour CPD-certified online course, available now at CESI COURSES – CESI Learning It includes a practical guidebook with a pre-interview checklist, a letter to employee template, and a reasonable adjustments checklist that you can refer to in real time.
For senior leaders and managers committed to building this understanding at a strategic level, the forthcoming CMI Level 7 Certificate in Autism-Informed Strategic Management and Leadership Practice opens for its first cohort in September 2026. You can register your interest here: Join the Waitlist – CESI Learning
Fair processes don’t happen by accident. They happen by design.
Dr Caroline Gorden is the founder and director of CESI Learning. Her book, Case Studies of Famous Trials and the Construction of Guilt and Innocence, is published by Bristol Policy Press. She can be found on LinkedIn at: https://www.linkedin.com/in/drcarolinegorden
References
Chandler, R., Russell, A., and Maras, K. (2018) ‘Compliance in autism: Does it occur and is it a problem?’, Research in Autism Spectrum Disorders, 56, pp. 10–21.
