AI ethics

Choosing an AI Ethics Speaker Who Can Move Beyond Principles

Evaluate AI ethics speakers through concrete dilemmas, stakeholder analysis, and practical accountability questions.

Top One Percent AI · Practical guidance with linked sources.

Fairness, transparency, and accountability are useful starting points for an AI ethics event. They are not a complete agenda. A strong session helps the audience recognize where these values create practical questions, whose interests may be overlooked, and who must act when a concern is identified.

For event planners, the selection challenge is to distinguish a speaker who can list principles from one who can help a particular audience reason responsibly about decisions.

Begin with audience responsibility

An ethics session for product teams should not be identical to one for a board or a human-resources audience. Product teams may need to examine design choices and evaluation methods. Leaders may need to clarify ownership and escalation. An audience without direct deployment authority may need to learn how to raise concerns effectively.

Describe the audience’s responsibilities in the brief. Avoid asking the speaker to cover every ethical issue in a single hour. Choose a small number of decisions that attendees can recognize without sharing confidential cases.

Use principles as a reference, not a conclusion

The OECD AI Principles address areas including fairness and privacy, transparency and explainability, robustness and safety, and accountability. They provide a shared vocabulary, but naming a principle does not settle how an organization should handle a specific situation.

In the selection interview, offer a hypothetical case: an automated prioritization tool improves overall throughput, but some affected people have difficulty challenging its output. Ask the candidate what additional information they would need before advising the organization.

A thoughtful response should explore who is affected, the consequences of mistakes, the available alternatives, and the practical route for review. Be cautious when a candidate turns an incomplete scenario into a confident verdict.

Look for treatment of competing interests

Ethical questions can involve tradeoffs rather than a single obvious answer. More explanation may help people understand a decision, but the form and detail of that explanation matter. More human review may help, but reviewers need time, information, and authority to change an outcome.

Ask the speaker to distinguish a stated safeguard from an effective one. “A human is involved” is a description of a workflow, not proof that the workflow protects people. The session should help attendees ask whether that person can detect a problem and intervene meaningfully.

Ask for evidence and boundaries

Request sources for factual claims and labels for hypothetical examples. If the speaker discusses law, ask how they keep jurisdiction and date clear. An ethics keynote should not drift into individualized legal advice or imply that a general checklist establishes compliance.

Also ask how the candidate handles audience disagreement. The goal is not to eliminate difficult questions through slogans. A facilitated discussion should allow uncertainty while keeping the conversation respectful and grounded.

Specify the takeaway

A useful output could be a short set of questions for project reviews: Who benefits? Who bears the downside? What can affected people understand or contest? Who owns follow-through? What evidence would show that a safeguard works?

Choose the speaker who can turn those questions into a repeatable conversation for your audience. The strongest result is not that attendees memorize ethical terminology. It is that they become more capable of identifying a consequential concern and bringing it to someone who can address it.