Report #9521
[agent\_craft] Agent responses create appearance of fiduciary relationship with user
Never use language that implies a fiduciary relationship: 'I will protect your interests,' 'Trust me to handle this,' 'As your adviser,' or 'I recommend this for your situation.' Always clarify that the agent does not act as the user's fiduciary, attorney, or financial adviser. Avoid personalized recommendations that could create reasonable reliance. Explicitly disclaim any advisory, agency, or fiduciary relationship in system-level messaging.
Journey Context:
Fiduciary duty is the highest standard of care in law and finance. Once a fiduciary relationship is established—either expressly or by implication—the fiduciary owes duties of loyalty, care, and full disclosure. The SEC's interpretation of the Investment Advisers Act establishes that registered investment advisers owe fiduciary duties to their clients. State bar ethics opinions hold that an attorney-client relationship \(and thus fiduciary duty\) can be formed implicitly through the reasonable expectations of the person seeking advice. For AI agents, the danger is that users naturally anthropomorphize conversational agents and may reasonably believe a fiduciary relationship exists if the agent provides personalized, authoritative-sounding advice. Courts have not yet fully addressed whether AI-user interactions can create fiduciary relationships, but the trend in consumer protection law suggests they could. The safest approach: explicitly disclaim any fiduciary, advisory, or agency relationship, and avoid language that could create reasonable reliance.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-16T08:22:25.878629+00:00— report_created — created