List confidentiality issues for generative tools in an outsourcing draft
Key facts
- Human involvement
- YellowClose supervision
- How AI is used
- List whether the draft says the vendor may train on inputs, who may see client information, and who must check output. Do not write a clause.
- What the lawyer must review
- Any clause that will be signed is yours. Opinion 512 treats client information entered into a generative tool as a confidentiality problem under Model Rule 1.6 when the tool trains on the input or can disclose it. The opinion also requires competence and a check of output. You decide the allocation of risk. The tool only shows what the draft already says.
- Written for / version
- United States versionUS
Who does what
| Role | What they do |
|---|---|
| Assistance | Potential support includes preparing preliminary structure or wording and identifying points for review. Confirm the tool’s capability and permitted scope. |
| Lawyer | Set the scope, supervise the work closely, verify every part relied on, resolve omissions and exceptions, and approve the exact output. |
Checklist
- Review operative clauses in the complete document, including definitions, cross-references, schedules and inconsistencies with the instructions.
- Confirm that the actual tool, feature and processing terms permit use of this information, including access, retention, training and disclosure conditions.
- Confirm the output against the client’s instructions and the complete relevant source material, including omissions or contradictions.
- Resolve exceptions and record the responsible lawyer’s approval of the exact output and intended use.