Is your firm's AI use compliant with ABA Opinion 512?

Do you understand, at a general level, how the AI tools your firm uses actually work — what they do with an input, and where they tend to be wrong?

Does someone at the firm independently verify every AI output — citations, quotations and figures — before it reaches a client or a court?

Do you know whether the AI tools you use train on, retain, or learn from the information you put into them?

If any tool you use does retain or train on inputs, have you obtained the client's informed consent before putting their confidential information into it?

Do you have a written data processing agreement — or equivalent contractual confidentiality terms — with each AI vendor holding client data?

Do you know where your client data physically sits, and what happens to it if you stop paying the vendor tomorrow?

Have you considered whether your clients should be told that AI is used in their matters, and made a deliberate decision either way?

Is your billing practice clear on what AI-related time you charge for — and are you not billing clients for the time spent learning a new tool?

Does your firm have a written AI policy that everyone — attorneys, paralegals and admin staff — has actually read?

Could you list, right now, every AI tool anyone at your firm uses on client information?