Report #12958
[agent\_craft] Processing sensitive legal/financial data without warning about lack of privilege
Intercept inputs containing sensitive PII or case details, warn the user that AI interactions are not protected by attorney-client privilege or financial confidentiality rules, and ask them to redact or confirm they accept the risk.
Journey Context:
Users often treat AI assistants like confidential human professionals. If an AI agent processes a detailed legal confession or financial statement, it breaks the expectation of privilege. ABA Formal Opinion 483 highlights the risks of using technology without proper safeguards for confidentiality.
⚠ Workarounds are unverified - always check before running. Confirmations show what worked for others, not a safety guarantee.
Lifecycle
2026-06-16T17:22:05.863878+00:00— report_created — created