Your associate just used AI to draft a motion that would have taken three hours. It took 40 minutes. The work product is solid, the client's matter moved forward, and you're staring at a timesheet wondering what number to write down.

This isn't a hypothetical. It's a billing question that every small practice using AI is already facing, whether they've thought about it carefully or not. And the ethical guidance is catching up fast.

In July 2024, the ABA issued Formal Opinion 512, its first comprehensive ethics guidance on lawyers' use of generative AI. The core message on fees was straightforward: lawyers billing hourly must bill for actual time spent, not the time a task would have taken without AI. Since then, state bars across the country have reinforced and expanded on that principle. Texas Opinion 705 (February 2025), Oregon Formal Opinion 2025-205, Florida's Opinion 24-1, and dozens of other state-level opinions have all landed in the same place.

The rule is clear. The compliance challenge is practical: how do you actually bill this way, prove you're doing it right, and keep your practice profitable in the process?

The Ethical Framework: What You Can and Can't Charge

The billing guidance across ABA and state bar opinions follows a consistent pattern. Understanding these principles isn't just good ethics; it's risk management for your practice.

Bill for Time Actually Spent

This is the foundational rule. If you're billing hourly and AI helps you complete a research memo in 45 minutes instead of three hours, you bill 45 minutes. As Texas Opinion 705 put it plainly, attorneys cannot charge hourly fees for time that was "saved" by using a generative AI program.

This applies even if the quality of the output is equal to or better than what manual work would have produced. The hourly billing model charges for time, not value. If you want to capture the value AI enables, that's a pricing model conversation (more on that below).

AI Tool Costs: Overhead vs. Billable Expense

The ABA drew an important distinction in Formal Opinion 512. If an AI tool functions more like general office equipment (think: AI spell-check in your word processor, or AI built into your standard legal research platform), it's overhead. You can't bill the client separately for it.

If, on the other hand, you use a third-party AI service that charges per use for a specific client matter, that cost can potentially be passed through as a disbursement, provided the client has agreed to the arrangement in advance and the charge reflects your actual cost.

The key word is "actual." You can't mark up AI tool costs. You can't add a surcharge. And if you can't separate a specific client's usage from your general subscription, the ethical path is to treat it as overhead.

Disclose Before You Bill

Multiple state opinions emphasize that transparency about AI use and billing needs to happen upfront, not after the invoice lands. Florida's Opinion 24-1 requires that lawyers notify clients if they intend to charge for AI tool usage. Oregon's opinion reinforces that billing practices must reflect the actual work performed and any AI-related costs must be clearly disclosed.

The safest approach: address AI use and its billing implications in your engagement letter. A brief paragraph explaining that your practice uses AI tools to improve efficiency, that hourly billing reflects actual time, and that any direct AI costs will be disclosed is far better than trying to explain it retroactively.

You Can't Bill for Learning to Use AI

This one catches people off guard. ABA Opinion 512 and several state opinions, including Oregon's, confirm that you generally cannot charge a client for time spent developing your competence with AI tools. Learning how to use ChatGPT or training on a new legal AI platform is your professional development. It's not billable unless a client specifically asks you to learn and use a particular tool for their matter.

The Compliance Gap That Should Worry You

Here's where the comfortable world of "I'll figure it out later" meets reality.

According to Clio's Legal Trends Report, 79% of legal professionals now use AI tools in some capacity — yet 44% of law firms have yet to implement formal governance policies, a gap the North Carolina Bar Association highlighted in its January 2026 AI guidance.

This isn't a problem for next year. Bar regulators are actively investigating billing practices right now. If a client questions an invoice or files a complaint, and your practice has no documented AI policy, no time tracking records that distinguish AI-assisted work, and no engagement letter language addressing AI use, you're exposed.

The ethics opinions are not vague about this. They're specific, published, and increasingly well-known to the clients who hire you.

Why Time Tracking Becomes Your Compliance Mechanism

Every piece of ethical guidance on AI billing points back to the same requirement: your billing must reflect actual time spent. That makes contemporaneous time tracking the single most important compliance tool in your practice.

Consider what you need to demonstrate if a billing dispute or ethics inquiry arises:

• That your hourly charges reflect hours actually worked, not hours estimated based on pre-AI timelines
• That you didn't bill the client for time spent learning how to use AI tools
• That any AI-related costs passed to the client are actual, documented expenses
• That the efficiency gains from AI benefited the client through lower bills, not your margin through inflated hours

Time tracking data is the evidence that makes all of this defensible. Without it, you're relying on memory and good intentions, neither of which holds up well in front of an ethics panel.

Track AI-Assisted Work Separately

This is a practical step that many practices overlook. When you use AI to assist with a task, your time entry should reflect that reality. Not in a way that's complicated or burdensome, but in a way that's clear.

For example, instead of logging "Legal research, 3.0 hours" when you actually spent 45 minutes using AI-assisted research and 30 minutes verifying the results, your time entries should reflect what happened: 45 minutes of AI-assisted research, 30 minutes reviewing and verifying output. That way, if anyone ever asks how a particular matter was billed, the answer is right there in the data.

This level of detail also helps you internally. Over time, you'll build a dataset showing exactly how much time AI saves on different task types, which is invaluable for making pricing decisions.

Capture Time in Real Time

The general principle of contemporaneous time capture is well established, and it becomes even more critical when AI is involved. Research cited by LeanLaw indicates that lawyers who wait until the end of the day to record their time lose 10 to 15% of billable hours. Wait until the end of the week and that figure can reach 25%.

When AI compresses what used to be a three-hour task into 40 minutes, the risk of misremembering the actual time is even higher. Did that contract review take 35 minutes or an hour? If you're reconstructing your day from memory, you might default to what the task "should have" taken rather than what it actually did. That's precisely the kind of billing inflation the ethics opinions warn against.

Track as you go. Every time. Especially on AI-assisted work.

Rethinking Your Pricing Model

The ethical framework on AI billing creates a real tension for hourly-billed practices. AI makes you faster, which under hourly billing means you earn less for the same (or better) output. Over time, this pressure will push many practices toward alternative pricing models.

Flat Fees Benefit from AI Efficiency

Under a flat fee arrangement, AI efficiency works in your favour. If you quoted $2,000 for a contract review and AI helps you complete it in half the time, you keep the full fee. The client got exactly what they paid for, at the price they agreed to.

But even here, the ABA has cautioned that it may be unreasonable to charge the same flat fee with AI as without it if the technology enables significantly faster completion. The answer isn't to ignore AI's impact on your pricing. It's to use time tracking data to set flat fees that reflect both the value you deliver and the time you invest, then adjust those prices as your AI-driven efficiency improves.

Hybrid and Value-Based Approaches

Many small practices are finding that a hybrid approach works well during this transition. Offer flat fees for well-defined, AI-friendly tasks (document drafting, standard research, contract review) where your efficiency gains are consistent and predictable. Keep hourly billing for complex, variable work where AI's role is less defined.

The common thread across every pricing model: you still need time tracking data. For hourly billing, it's your compliance mechanism. For flat fees, it's how you measure profitability and know whether your prices are right. For hybrid models, it's how you decide which matters belong in which category.

Building Your AI Billing Policy: A Practical Checklist

You don't need a 50-page governance document. You need a clear, written policy that covers the essentials. Here's what to include:

Engagement Letter Language

Add a paragraph to your standard engagement letter that covers AI use, how AI-assisted work is billed (actual time only for hourly matters), whether AI tool costs may be passed through, and how the client can ask about AI use on their matter.

Internal Time Tracking Guidelines

Establish clear guidelines for your team: track all AI-assisted work with enough detail to distinguish it from manual work, capture time contemporaneously (not reconstructed later), and note when AI tools are used on specific tasks.

Tool Approval and Cost Documentation

Maintain a list of approved AI tools, their costs, and whether those costs are treated as overhead or client-specific expenses. This documentation is both good governance and good evidence if a billing question ever arises.

Quarterly Review

Review your AI-assisted time data quarterly. Look at which tasks AI is accelerating, whether your pricing reflects the new reality, and whether your time tracking practices are consistent across the team.

Where MinuteDock Fits In

The ethical requirements around AI billing all converge on one point: you need accurate, contemporaneous, matter-level time tracking that you can actually use for compliance and pricing decisions.

MinuteDock makes it easy to capture time in real time, whether you're billing hourly or tracking for flat fee profitability analysis. Assign Time Entries to specific Contacts, Projects, and Tasks so that your AI-assisted work is clearly documented and organised. Add notes to each entry to record what tools were used and what review was completed.

Set budgets for flat fee matters to monitor your effective hourly rate in real time. When AI accelerates your delivery, you'll see it reflected in your profitability data immediately, giving you the insight to adjust pricing with confidence rather than guesswork.

And because MinuteDock syncs with your accounting platform (Xero, QuickBooks, and others), your time data flows directly into your billing and financial reporting. No manual reconciliation, no data gaps, just a clean audit trail from tracked time to invoice.

In an environment where your billing practices need to be defensible, your time tracking data is your best evidence. Make sure it's accurate, detailed, and captured as the work happens.

The Bottom Line

AI is already changing how legal work gets done. The ethics rules aren't trying to stop that. They're asking you to be honest about it, bill for what you actually do, and give your clients the benefit of the efficiency you've gained.

The practices that handle this well won't just avoid ethics complaints. They'll build trust with clients who appreciate transparency, make smarter pricing decisions backed by real data, and position themselves as the kind of practice that clients actively want to work with.

The practices that wing it, billing by feel and hoping nobody asks, are taking a risk that grows every time a new state bar opinion lands.

Your time tracking data is what makes the difference. Start capturing it properly now, before the question comes up.

Try MinuteDock free and build the billing compliance foundation your practice needs.