Build a matter upload checklist before any file goes into a tool

Key facts

Human involvement: YellowClose supervision
How AI is used: For the categories the lawyer names, list whether the material is information relating to a representation, work product, or protected health information, and the restriction the lawyer has already marked. Do not upload anything.
What the lawyer must review: Comment 18 to Model Rule 1.6 requires competent safeguards against unauthorised access and inadvertent disclosure. Formal Opinion 512 treats input of information relating to the representation into a self-learning tool as a confidentiality problem unless the client gives informed consent or the disclosure is impliedly authorised. A covered entity or business associate may not use or disclose protected health information except as 45 CFR § 164.502 permits or requires. You decide the category and whether the tool is allowed to see it.
Written for / version: United States versionUS

Who does what

Who does what
Role	What they do
Assistance	Consider narrowly defined preparation support after confirming 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.

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.

Forms and precedents for this task
Form or precedent	Kind	AI use	Summary
Protocol before using a self-learning generative AI tool (United States)	Checklist	AI use Yellow	A check before anyone puts information relating to a representation into a generative AI tool that may learn from inputs or reveal them in later output, including to someone behind an ethical wall.
Staff undertaking on unapproved generative AI tools (United States)	Letter	AI use Yellow	A short undertaking by a lawyer or staff member not to put firm or client information into a generative AI tool the firm has not approved, and to tell a supervisor if that happens.
Check before using a hypothetical in a generative AI tool (United States)	Checklist	AI use Yellow	A check that a prompt is a true hypothetical under Rule 1.6 Comment [4]: no reasonable likelihood that the tool, or a later reader of the output, could identify the client or the situation.