Should You Bill AI Time to Clients? The 2026 Guide to Ethical AI Pricing

The Question Every AI-Using Lawyer Eventually Faces

You spend 30 minutes on a research project that used to take 4 hours. You bill the client. How much do you charge?

This question has tied lawyers in knots for two years. State bar opinions are scattered. ABA guidance is still evolving. Clients are starting to ask pointed questions.

This article gives you the 2026 framework: what you can ethically bill, what you can't, and how to structure your engagement letters and time entries to stay safely on the right side of the line.

The Bedrock Rule: ABA Model Rule 1.5

Model Rule 1.5(a) prohibits "unreasonable" fees. Comments and ethics opinions have repeatedly held that fees must reflect the actual value delivered and the actual time spent. Both factors matter.

For AI, this creates a clean test: you can charge for what AI helps you deliver, but you cannot charge twice for the same work.

For the broader ethics framework (Rules 1.1, 1.6, and 5.3), see AI ethics for lawyers.

The Three Buckets of AI Time

For billing purposes, AI use in legal work falls into three buckets:

Bucket 1: AI as background productivity tool. You used spell-check, your word processor, your calendar, AI-powered email triage. These are background tools that you don't separately bill - and you don't separately bill AI when used this way.

Bucket 2: AI doing a discrete task you supervise. You ask AI to draft a demand letter. You review and revise. You bill for the time you actually spent on the matter (supervision, review, judgment).

Bucket 3: AI replacing what was previously a billable human task. You used to assign a junior associate 4 hours of document review. Now AI does it. The cost shifts - and the billing approach shifts with it.

The hardest cases are in Bucket 3. The rest of this article focuses there.

The Honest Time Rule

The cleanest rule, endorsed by the ABA's 2024 Formal Opinion 512 and most state bars that have weighed in: bill for the time you actually spend, not the time the task used to take.

If AI does in 20 minutes what used to take 4 hours, you bill for 20 minutes (or whatever time you actually spent supervising, reviewing, and exercising legal judgment).

You cannot bill for the "saved time." You also generally cannot bill the AI's "time" - it's not your time.

This is the bright-line answer. Memorize it.

When You Can Charge More Than Time

The big asterisk: hourly billing is not the only model. The honest-time rule doesn't bind alternative fee arrangements.

If you use:

• Flat fees
• Value-based fees
• Capped fees
• Project-based fees
• Subscription / retainer arrangements

...you can structure the fee around the value of the work, not the hours spent. AI then becomes an efficiency multiplier within your fee structure rather than a reason to discount your bill.

This is the strategic move. Firms that capture the most AI value in 2026 are shifting away from pure hourly billing into models where AI efficiency translates to firm margin.

What About AI Tool Costs?

You can generally pass through AI tool costs to clients as costs, the same way you pass through Westlaw or court filing fees - but only if your engagement letter authorizes it.

The recommended approach:

• Charge AI tool costs as a flat per-matter "technology fee" or "research and AI access fee."
• Disclose this in your engagement letter.
• Avoid line-itemizing individual AI tool charges in client invoices unless the engagement specifically calls for it.

Some firms have started charging an "AI technology fee" of $50 to $150/matter for routine matters and more for complex ones. This is reasonable, transparent, and accepted by most clients.

What You Must Disclose

ABA Formal Opinion 512 and most state bar opinions have converged on the following disclosures:

1. To clients: Disclose AI use in engagement letters when AI will substantively touch client work. Update your standard engagement letter language now.

1. To clients (billing): If you charge a technology fee or pass through AI costs, disclose this in the engagement letter and on invoices.

1. To courts: Some jurisdictions now require disclosure of substantive generative AI use in filings. Check local rules in every jurisdiction where you appear.

1. To opposing counsel: Generally not required absent specific rules. Some judges encourage disclosure during meet-and-confer on production protocols (see AI for document review and eDiscovery).

Sample Engagement Letter Language

Add this (or your jurisdiction-specific equivalent) to engagement letters today:

"Use of Artificial Intelligence. The firm uses generative artificial intelligence tools as part of its practice. AI tools may assist with research, drafting, document review, and similar tasks. All AI-assisted work is reviewed by a licensed attorney. AI tool costs may be included in the firm's flat technology fee of $[amount] per matter, or passed through as a disbursement. The firm does not bill for time savings achieved by AI; the firm bills only for actual attorney and staff time spent on the matter."

Have your malpractice insurer and ethics counsel approve specific language for your jurisdiction.

The Hourly Lawyer's Dilemma

If your practice is built on hourly billing, AI creates a structural problem: the better you get at AI, the fewer hours you bill. A 4-hour memo becomes 30 minutes. Your revenue per matter falls.

There are three rational responses:

1. Use the saved hours to take on more matters. Same revenue, more volume.
2. Shift selectively to flat fees and value pricing where AI gives the most leverage.
3. Raise your hourly rate to reflect the increased value delivered per hour.

The wrong response is to over-bill AI work to maintain hourly revenue. That's the sanctions case waiting to happen.

The Flat-Fee Renaissance

Some practice areas are tailor-made for AI-enabled flat fees:

• Estate planning packages
• Business formation packages
• Standard contract review
• Routine demand letters (see draft demand letters 5x faster with AI)
• Trademark filings
• Family law document preparation
• Routine immigration filings

The pattern: predictable scope, high AI leverage, client preference for fee certainty.

Firms that move these practice areas to flat fees and let AI compress the work hours can dramatically increase margins. Many of our clients have moved 30 to 40% of their work to flat fees in the last 18 months. See our case studies in the dashboard every managing partner wants.

Common Mistakes to Avoid

1. Double-billing. You cannot bill 4 hours of "research" plus AI tool fees when actual time was 30 minutes.
2. Hidden AI fees. Don't surprise clients with line-item AI charges that weren't disclosed.
3. Padding time entries. Don't write up to "what it would have taken."
4. Charging for AI-only time. The AI's processing isn't your billable time.
5. Failing to disclose. Always disclose AI use in engagement letters.

A Self-Audit Exercise

Spend 30 minutes this week running this audit:

1. List the matters you've billed in the last 30 days.
2. For each matter, note: did you use AI substantively? In what tasks?
3. Compare the time billed to what you'd defend in front of a fee review.
4. Identify any matters where billing might not pass scrutiny.
5. Adjust going forward.

If even one matter feels uncomfortable, your billing practices need adjustment.

Where This Is Heading

By 2027 to 2028, we expect:

• Mandatory AI disclosure in engagement letters in most jurisdictions
• Standardized AI technology fee structures
• Increased client pressure for value-based pricing
• Fee disputes specifically over AI billing in fee arbitrations and litigation

The firms that update their billing models now will avoid all of that. The ones that don't will be reactive, often expensively.

Where to Get Help

If you want to update your engagement letters, billing policies, and pricing structure for the AI era, our AI Operating System service includes ethics-compliant billing infrastructure as part of every implementation. Or book a consult to discuss your specific practice.

AI is the biggest business model change in legal services in a generation. The lawyers who think about billing first - not last - will win this transition.