In conventional research, participation is sometimes a methodological choice:   it produces richer data. In Action Research, participation is an ethical commitment.

The reasons are multiple and interconnected.

Epistemic justice: the people most affected by AI systems have knowledge that the people building those systems do not. They know how decisions feel from the inside. They know which aspects of the system’s behaviour matter most. They know the trade-offs that look acceptable on paper but are experienced as harmful in practice. To conduct research about AI impacts without including these people is not just methodologically limited:   it is ethically problematic.

Self-determination: people have the right to participate in decisions that affect them. An AI system that makes consequential decisions about people’s access to credit, healthcare, employment, or justice is making decisions that profoundly affect those people’s lives. Involving them in research about the system’s ethics is a minimal expression of respect for their autonomy.

Better outcomes: participatory research consistently produces more accurate diagnoses and more viable solutions. Solutions designed with those affected are more likely to address real needs, avoid unintended harms, and be accepted and used.

Trust: in a field where AI has frequently been experienced as something done to people rather than with them, genuine participation in research builds the trust that ethical AI ultimately requires.

In practice, participation in ethical AI Action Research means:
– Involving community members in framing the research questions, not just answering them
– Creating accessible processes that do not require technical expertise to participate
– Sharing findings back with participants before they are published or acted upon
– Ensuring that participation is compensated and genuinely voluntary
– Being transparent about what will and will not change as a result of the research

Participation is not easy. It takes more time. It complicates neat narratives. It surfaces uncomfortable truths. But it is the foundation of ethical AI research that deserves the name.

Reflection question: Who is currently excluded from research and decisions about AI systems that affect them in your organisation or community? What would genuine participation look like for them?