Artificial Intelligence (AI) continues to revolutionize how legal professionals conduct research, analyze data, and produce documents. For solo and small law firms in particular, AI seems to promise efficiency and affordability. But the convenience of free AI tools can come at a steep ethical and professional cost—especially when dealing with confidential client information such as medical records, police reports, or financial disclosures.

This article highlights the specific risks of free AI platforms, summarizes the North Carolina State Bar ethics opinion on AI, and offers practical tips for safely integrating AI into your practice.

What the NC State Bar Says About AI Use

In 2024 Formal Ethics Opinion 1, the North Carolina State Bar made clear: lawyers are permitted to use AI in their practices—but only if they do so competently, securely, and with proper supervision. Key takeaways include:

• AI must not compromise client confidentiality: Lawyers are required under Rule 1.6(c) to make reasonable efforts to prevent unauthorized access to client information. That includes ensuring the AI platform won’t use client data to train its model.
• Competence includes technological understanding: Per Rule 1.1, lawyers must stay abreast of technological changes—including understanding the benefits and risks of AI tools used in legal practice.
• You can’t outsource responsibility: Whether work is done by a paralegal, summer associate, or AI program, the lawyer must independently review and take full responsibility for all outputs used in client matters.
• Client consent may be required: If a lawyer uses AI in substantive legal tasks (e.g., drafting estate plans, contracts, or pleadings), and especially if the tool affects fees or billing, the client must be informed under Rule 1.4.
• Avoid free public AI tools for client-specific data: The opinion cautions lawyers against inputting sensitive information into publicly accessible AI platforms that may store, retain, or use submitted data for model training.

High-Risk Areas for Confidentiality Breaches

Some legal practice areas are especially vulnerable when using unsecured AI tools:

• Personal Injury & Medical Malpractice: Summarizing client medical records in free tools may violate HIPAA.
• Family Law: Custody evaluations, financial affidavits, and therapy notes must remain confidential.
• Criminal Defense: Sharing police reports or client statements in AI platforms could undermine privilege.
• Elder Law & Estate Planning: Drafting wills or powers of attorney through AI requires safeguards against data leakage.

Red Flags in Free or Consumer-Grade AI Tools

Ask the following before using any AI platform:

• Is the platform trained on user-uploaded data?
• Are there binding terms about confidentiality and data retention?
• Is data encrypted at rest and in transit?
• Can you delete client data from the system permanently?
• Is this platform created for legal use—or for general consumers?

If the answer is uncertain, do not use that platform with client information.

Best Practices: How to Use AI Responsibly in Your Firm

1. Choose the Right AI Platform

Pick vendors who design tools specifically for legal professionals and who offer:

• Written confidentiality and non-disclosure agreements
• Data residency and encryption
• Opt-out provisions for model training
• Clear client data ownership and deletion policies

Evaluate vendors by reviewing their company reputation, data security protocols, and contract provisions for how data is stored and destroyed if services end.

2. Vet Every AI Tool Like a Vendor

Think of AI platforms like any third-party contractor. You are obligated to ensure they operate in a way that’s compatible with your ethical duties. This includes:

• Periodic audits
• Security consultations
• Ongoing training for your team

3. Maintain a “Human in the Loop”

Inquiry 4 of the AI opinion asks “if a lawyer signs a pleading based on information generated  from AI, is there variation from traditional or existing ethical obligations and expectations placed on lawyers signing pleadings absent AI involvement?” The opinion response is “no, a lawyer may not abrogate their responsibilities under the Rules of Professional Conduct by relying on AI”. 

You must always review the AI’s output. AI hallucinations (inaccuracies) are still a real risk. Use the correct AI tools – for example, don’t trust legal citations to ChatGPT or Microsoft CoPilot, rely on the legal research tools which have their AI incorporated into the products for brief writing or legal research.

4. Be Transparent About Fees

The State Bar prohibits the lawyer for billing a client for three hours of work when only one hour of work was completed. While AI will provide efficiencies, the lawyer’s billing must be accurate, honest and not clearly excessive. Most law firms need to engage in discussion to determine how to bill for efficiencies that result from using AI tools. Flat fees or itemized charges for AI usage are allowed, provided they are transparent and not excessive.

5. Avoid Inputting Confidential Information Into Free Tools

Unless the tool offers specific enterprise safeguards or is configured for legal compliance, don’t risk uploading client files. This includes tools like ChatGPT, unless using a business or enterprise version that allows full control over data privacy.

Bottom Line: Ethics Over Efficiency

AI can support faster research, more consistent drafting, and improved client service. But convenience should never come at the cost of compliance—or client trust. By choosing tools carefully, training staff, and maintaining independent professional judgment, even solo and small firm lawyers can benefit from AI while staying fully within ethical bounds.