The legal profession is not new to ethical questions raised by technology. Email, cloud storage, and outsourcing have each prompted bar opinions and practice changes. AI is the next chapter — and it deserves more careful thought than most firms are giving it.

‍

The duties of competence, confidentiality, supervision, and candor apply to AI tools the same way they apply to any other technology. Most state bars have published guidance affirming this, and the ABA's Formal Opinion on generative AI is now the baseline reference.

‍

The Existing Framework

What the rules require is largely what good practice would require anyway: understand what the tool does, protect client data, supervise the output, and don't claim work as your own that you haven't reviewed.

‍

‍

Disclosure to Clients

Whether to disclose AI use to clients is one of the most-debated questions in the profession. We come down firmly on the side of disclosure — not because the rules currently require it in every case, but because we'd rather build a practice on transparency than hope a client doesn't ask.

Our engagement letters describe how we use AI, where we don't, and what protections are in place. Clients can opt out of specific uses. Most don't.

‍

The Hard Cases

There are genuinely difficult ethical questions where the rules don't yet provide clear answers. Can a firm hold itself out as having particular expertise that is largely embodied in its tools? How should AI-assisted work be billed? What does "supervision" mean when the model produces a draft no human has read in full?

We don't claim to have settled answers. We do claim to have a process for working through them — one that involves our ethics committee, outside counsel, and ongoing dialogue with the bar.

‍

‍

The Standard We Set

The standard we set for ourselves is higher than the floor the rules require. Every output that leaves BauHaus has been reviewed by a human attorney who is accountable for it. We document our AI use. We disclose it. We invite scrutiny.

If the profession is going to retain public trust through this transition, that's the standard the field needs.

‍