Skip to main content
    Back to The Human Advantage
    Adaptation
    6 min read8 April 2026

    Whether people can direct, combine and check AI agents in real work.

    Strategic Oversight and Ethical Governance in AI Orchestration

    AI capability has moved faster than accountability. Here is what ethical governance means day to day, a checklist for AI-supported decisions and why oversight belongs with leaders.

    Gemma George

    Co-Director & Client Service Manager

    Ethical governance is the sixth of the six orchestration capabilities. AI capability has moved faster than the structures that hold it to account.

    Information about models is getting thinner. Stanford's 2026 AI Index reports that its foundation model transparency score dropped from 58 to 40.

    Trust is fragile too. The CIPD's 2026 Good Work Index found 45% of employees using generative AI do not trust its output to be accurate.

    What ethical governance means in practice

    Floridi and colleagues' AI4People framework draws together core principles for AI ethics. In daily work they become five questions.

    Accountability. A named person answers for every AI-supported decision. "The system recommended it" is never enough.

    Transparency. People affected can understand how a decision was reached and what role AI played.

    Fairness. AI trained on past data can repeat past bias. Check whether it disadvantages particular groups.

    Privacy. Data use matches the law and people's reasonable expectations.

    Human dignity. People are treated as people, with a route to human review.

    Build this capability in your organisation

    Growth Performance designs bespoke programmes that turn insight into measurable behaviour change.

    Explore Adaptation

    An ethical pre-flight checklist

    Before an AI-supported decision, ask:

    • Is a named person responsible, and can they explain the decision?
    • Could you explain AI's role to someone affected?
    • Might this system advantage or disadvantage particular groups?
    • Does it use personal data as people would expect?
    • Would you be comfortable if those affected saw how the decision was made?
    • Is AI the right tool here, or simply the available one?

    The regulatory picture

    The EU AI Act sets rules by risk. Some practices are banned outright, and high-risk systems need documentation and human oversight.

    Its obligations are phasing in. Bans on some practices and a duty to build staff AI literacy have applied since February 2025.

    The direction is clear wherever you operate. AI governance is moving from good practice to legal duty.

    A leadership responsibility

    Legal and data teams provide essential expertise. Managers who run AI-supported work see the risks first.

    Build ethical review into workflow design. Make raising concerns normal, and record how significant decisions were made.

    Making it safe to question AI use builds trust. That is part of Adaptation, where people direct, combine and check AI agents in real work.

    Try this

    For each AI tool your team uses, name the person accountable if it causes harm. Where nobody is named, agree an owner this week.

    Then take the free AI Change Leadership Readiness assessment.


    Want to explore these ideas further?

    Tell us about your people and we will shape options around them.

    Get options for your team

    Trusted by McDonald's, Nissan, the NHS and Munich Re. 97% average satisfaction across 500+ programmes and workshops.Source: Growth Performance internal research data and surveys, June 2016 to June 2026.