AI Document Drafting and Legal Ethics

As AI document drafting becomes standard practice in US law firms, ethical obligations are top of mind for managing partners and ethics committees. The good news: existing rules of professional conduct already provide a framework. The challenge is translating abstract duties into concrete workflows.

Bar associations across multiple states have issued guidance confirming that AI use is permissible — and expected — under competence rules, provided attorneys maintain supervision and verify output. The ethical standard is not whether you use AI, but how you govern it.

Four ethical pillars for AI drafting

1. Competence (Model Rule 1.1)

Attorneys must understand the capabilities and limitations of AI tools they use. This means training on how drafts are generated, what data the system accesses, and where errors are most likely. Competence also requires knowing when AI is inappropriate — novel legal questions, highly regulated filings and sensitive negotiations may require traditional drafting approaches.

2. Confidentiality (Model Rule 1.6)

Client data used in AI drafting must remain confidential. Firms using Microsoft 365 Copilot benefit from data remaining within their tenant with existing permission controls. Avoid consumer-grade AI tools that may use inputs for model training. Review vendor data processing agreements carefully.

3. Supervision (Model Rule 5.3)

AI output must be reviewed by a licensed attorney before it reaches a client or court. Treat AI as a non-lawyer assistant whose work requires partner or senior associate review. Document your review process and maintain audit trails.

4. Candour and honesty

Attorneys must not misrepresent AI-generated work as entirely human-authored where disclosure is required. Some courts now require certification that filings were reviewed for accuracy regardless of drafting method. Check local rules and update your filing procedures accordingly.

Practical guardrails for your firm

• Adopt a written AI usage policy covering approved tools, prohibited uses and review requirements.
• Use firm-specific templates and verified precedent as grounding sources — not open-ended prompts.
• Require attorney sign-off on every AI-assisted document before client delivery or court filing.
• Train all fee-earners on AI limitations, including hallucination risks and citation errors.
• Conduct periodic audits of AI-generated work product for quality and compliance.

Akili Global helps US law firms implement AI drafting workflows with built-in ethical guardrails on Microsoft 365. Ethical AI use is not a constraint on efficiency — it is the foundation that makes efficiency sustainable.