Will AI make our workplace more inclusive automatically?

No. AI will amplify the culture it enters. If your organisation already has bias in its processes, language or decisions, AI will accelerate that — not correct it. Intentional governance, human oversight, and inclusive prompting are what turn AI from a risk into a genuine tool for belonging.

How can AI support neurodivergent or disabled colleagues?

Thoughtfully used, AI can help neurodivergent colleagues structure written communication, prepare for difficult conversations, or draft emails when executive function is low. It can generate transcripts, summarise long documents, and produce alternative formats. The key word is "support" — it removes friction, not the person's agency.

What does responsible AI adoption look like in practice?

It starts with a framework: human judgement plus critical thinking plus inclusive governance. That means human review before AI output reaches sensitive decisions, clear escalation routes when something feels wrong, psychological safety so people can raise concerns without fear, and ongoing reflection on whether AI tools are working for everyone — not just the majority.

Using AI for Good: From Risk to Belonging

Using AI for good at work means moving beyond the question of whether AI is risky and asking instead: how do we use it intentionally, with inclusive governance and human judgement at the centre? AI can support belonging — helping neurodivergent colleagues structure communication, improving accessibility, and giving managers a reflection tool for their language — but only when the organisation it enters has already committed to building an inclusive culture.

The question organisations are asking wrong

Most workplace conversations about AI start with risk: hallucinations, bias in hiring algorithms, data privacy, job displacement. These are real concerns and they deserve attention. But stopping at risk means organisations either ban AI out of fear or adopt it uncritically out of enthusiasm, and neither position serves the people doing the work.

The better question is: how do we use AI with intention? That means asking who benefits, who is left out, and what human judgement needs to remain in the loop before AI output shapes a decision about a person.

How AI can actively support inclusion

When used thoughtfully, AI removes friction that disproportionately burdens people who are already navigating more barriers. Some practical examples:

Neurodivergent colleagues can use AI to help structure written communication, draft emails when executive function is depleted, or prepare for conversations that feel cognitively demanding.
Accessibility improves when AI generates transcripts, summarises long documents, converts formats, or produces plain-English versions of complex policy.
Managers can use AI as a reflection partner — checking their language in performance reviews, testing whether feedback is specific and fair, or thinking through how to approach a sensitive conversation before having it.
Learning access opens up when people can ask AI to explain a concept in their own terms, at their own pace, without the social risk of asking a question in a room.
Difficult conversations become less daunting when someone can rehearse their thinking with an AI coach before raising a concern with HR or a line manager.

The framework: human judgement + critical thinking + inclusive governance

AI as coach, augmenter and reflection partner is genuinely useful. But it requires a framework. I work with organisations around three pillars:

Human judgement. AI output is a starting point, not a conclusion. Anywhere a decision touches a person — hiring, performance, reasonable adjustments, disciplinary — a human must review and own the outcome.
Critical thinking. Inclusive prompting matters: the questions you ask AI shape the answers you get. Teaching people to interrogate AI output — to ask "whose perspective is missing here?" — is as important as teaching them to use the tools.
Inclusive governance. Clear escalation routes when something feels wrong. Psychological safety so people can raise concerns without fear of being seen as obstructive. Policies that are reviewed for who they exclude, not just what they permit.

The amplification problem

My signature observation on this is blunt: AI will not automatically create a more inclusive workplace — it will amplify the culture it enters. If your recruitment process already disadvantages certain candidates, an AI screening tool will do that faster and at greater scale. If your performance review language already reflects who the manager is most comfortable with, AI-assisted templates will embed that pattern.

This is not an argument against AI. It is an argument for doing the cultural work first — or at least alongside — the technology adoption. The future of belonging and AI is not determined by the tools themselves; it is determined by the intentions, governance and humility of the people deploying them.

Psychological safety and experimentation

One of the less-discussed inclusion risks around AI is what happens when people are afraid to experiment with it. If the culture signals that using AI is cheating, or that getting it wrong is a disciplinary matter, the people who will quietly use it anyway are those with the most confidence and the least to lose. Everyone else falls further behind.

Building psychological safety around AI experimentation — naming it as a learning curve, not a competency test — is itself an inclusion intervention. It ensures the benefits of AI are distributed, not hoarded by those already advantaged.

From risk to intentional practice

The shift I encourage organisations to make is from asking "is AI risky?" to asking "how do we adopt it responsibly?" That means:

Auditing existing processes for bias before AI is applied to them.
Defining which decisions require human sign-off, always, without exception.
Training people in inclusive prompting — better questions produce better outputs.
Creating visible escalation routes and making it easy to flag concerns.
Reviewing AI tools through the lens of who benefits and who is systematically disadvantaged by them.

Explore how this connects to the broader topic of AI, belonging, and the future of work, or hear Joanne explore these ideas in depth on the Inclusion Bites podcast.

Using AI for Good: from risk to belonging