Question

A financial services firm deploys a Copilot Studio agent that recommends loan restructuring options to relationship managers. Regulators require that a human remain accountable for every credit decision, and the firm must be able to demonstrate this to auditors. Which design approach best aligns with the responsible AI principle of accountability while satisfying the regulatory requirement?

Accepted Answer

Configure the agent to present recommendations that a relationship manager must explicitly review and approve before any restructuring action is executed, and log each human decision with the agent's supporting rationale.

Answer

Enable autonomous orchestration so the agent applies the highest-confidence restructuring option automatically, and rely on the model's content safety filters to prevent inappropriate decisions.

Answer

Add a transparency note to the agent interface disclosing that it uses AI, and allow the agent to finalize decisions without further human involvement.

Answer

Route all agent recommendations to a scheduled monthly review board that samples 5 percent of executed decisions for retrospective validation.