# Can You Bill for AI-Assisted Legal Work?

> Can you bill clients for AI-assisted legal work? A plain walkthrough under ABA Opinion 512 and Model Rule 1.5, plus documentation that keeps a bill defensible.

[Home](https://rankshieldlegal.com/) / [Blog](https://rankshieldlegal.com/blog/) / Legal AI AI ethics & billing
# Can You Bill for AI-Assisted Work? The Rules and How to Document It
Sometimes, but not the way many firms assume. You can bill for AI-assisted work when the time was actually spent and the total fee stays reasonable, but you generally cannot bill a client for the hours a task would have taken before AI, and you generally cannot bill for learning a general-purpose tool. This is informational, not legal advice.

By [Jamie Kloncz](https://rankshieldlegal.com/about/), Founder, RankShield ** 22 min read ** Published July 3, 2026

The honest short answer is that it depends on what you are billing for, and the dividing line is narrower than a lot of firms assume. You can bill for AI-assisted work when the time was genuinely spent and the overall fee remains reasonable, but ABA Formal Opinion 512 makes clear that you generally cannot bill a client for time you did not actually spend, and you generally cannot pass along the cost of learning to use a general-purpose AI tool [[1]](#ref-1). The question is not whether AI touched the work. It is whether the bill reflects reasonable fees for effort that was real.
This piece is written by a founder who builds verification tools for law firms, not by an attorney, and it is general information rather than legal advice on any specific bill or fee arrangement. Billing is governed by Model Rule 1.5 and its state-adopted equivalents, and the reasonableness standard is fact-specific, so the sections below lay out how Opinion 512 applies that rule to AI, where the common traps are, and how to document AI-assisted work so a bill can be defended rather than merely asserted. Where the guidance varies by jurisdiction, this article says so and points you back to your own counsel and your licensing authority's rules.

## The short answer is that it depends on whether the time was real
Opinion 512 applies Model Rule 1.5 to AI. You can bill for AI-assisted work, but you cannot bill for time you did not spend, and the total fee has to stay reasonable no matter how much AI sped things up.
There is no rule that says AI-assisted work is unbillable, and there is no rule that says using AI lets you bill the way you always did. ABA Formal Opinion 512, issued in July 2024, applies the existing reasonable-fees duty in Model Rule 1.5 to generative AI, and the core of that application is simple to state even though it is fact-specific to apply: a lawyer may not bill a client for time not actually spent, and the total fee must remain reasonable regardless of how AI changed the work [[1]](#ref-1) [[2]](#ref-2). What you are selling is competent legal work at a reasonable fee, not a fixed number of hours.
That framing matters because it separates two things that firms tend to blur. The first is whether AI helped do the work, which it plainly can. The second is what the client is charged, which is governed by reasonableness and by the requirement that billed time correspond to time actually spent. A tool that drafts a first-pass research memo in ten minutes did not make the work worthless, but it also did not make it worth the two hours the task used to take. On an hourly bill, the client pays for the ten minutes plus whatever review and revision the lawyer actually performed.
So the useful move is not to look for permission to bill for AI as a category. It is to learn which billing practices Opinion 512 treats as improper, which ones stay within Rule 1.5, and how to keep records that show the difference. The rest of this article does that, keeps the reasonableness question in front throughout, and ends where a responsible version of this has to end, with your own counsel and your jurisdiction's adopted version of Rule 1.5.

## What ABA Formal Opinion 512 says about billing and fees
Opinion 512 addresses several professional duties that generative AI use implicates, and one of them is the reasonable-fees duty under Model Rule 1.5 [[1]](#ref-1). The opinion does not create a new billing rule for AI. It reads AI into the rule that already governs every fee a lawyer charges, which is that fees and expenses must be reasonable [[2]](#ref-2). The effect is to hold AI-assisted billing to the same standard as any other billing, and then to work through what that standard means when a tool changes how long a task takes.
On fees specifically, the opinion's guidance is conditional and grounded in reasonableness rather than in a flat prohibition. A lawyer who uses AI to complete a task faster may bill for the time the lawyer actually spends, but may not bill as though the pre-AI process was used when it was not [[1]](#ref-1). The reasonableness requirement in Rule 1.5 remains the backstop for every fee arrangement, hourly or otherwise, so even where a fee is not tied to hours, the amount still has to be reasonable in light of the work performed [[2]](#ref-2).
Reading the opinion as a ban on billing for AI misses how it works, and so does reading it as a green light. It ties fee reasonableness to competence and communication, so that a lawyer who uses these tools understands them well enough to bill honestly for the work and, where relevant, explains AI-related costs to the client [[1]](#ref-1). Those obligations travel together. A defensible AI bill is one where the time is real, the fee is reasonable, and the client understood the basis for what they were charged.

- **Reasonable fees** Under Rule 1.5, every fee and expense must be reasonable, and Opinion 512 applies that duty to AI-assisted work without carving out an exception [[1]](#ref-1) [[2]](#ref-2).
- **Actual time** A lawyer may bill for time actually spent using and reviewing AI output, but not for time the pre-AI process would have taken [[1]](#ref-1).
- **Communication** Under Rule 1.4, where AI affects the cost of the representation, that is something a lawyer may need to explain to the client [[1]](#ref-1).
Opinion 512 does not invent a separate AI billing rule. It applies Model Rule 1.5's reasonableness standard to AI-assisted work, and leaves the specific determination fact-specific.

Source: ABA Formal Op. 512; ABA Model Rule 1.5 Download SVG

## You cannot bill for time you did not actually spend
If AI cuts a task from two hours to twenty minutes, the hourly bill is the twenty minutes plus real review time. The output being valuable does not turn unspent time into billable time.
The clearest line Opinion 512 draws on billing is also the oldest one in fee ethics, applied to a new fact pattern. A lawyer may not bill a client for time that was not actually spent, and AI does not change that [[1]](#ref-1). If a research task that once took two hours now takes twenty minutes because a tool produced a usable draft, the hourly bill reflects the twenty minutes plus whatever verification and revision the lawyer genuinely performed. Billing the old two hours because that is what the task "is worth" is billing for time that did not happen.
This is worth stating plainly because the temptation runs the other way. A firm that has priced a task at a certain number of hours for years can feel that AI made the same work product cheaper to produce but no less valuable, and reason from there to keeping the hours the same. Opinion 512 does not accept that reasoning for hourly billing. The value of the output does not convert unspent time into billable time. What the lawyer can bill is the time spent directing the tool, checking its output, correcting it, and doing the parts of the work AI did not do.
The verification time is real, and it is often substantial, which is part of why honest AI billing is not the same as billing nothing. Confirming that AI-generated research is accurate, that citations exist and say what the draft claims, and that the analysis fits the matter takes lawyer time that is properly billable. The discipline is to bill that actual time rather than a legacy estimate. For why the verification step is not optional, see [our guide to building a law firm AI policy](https://rankshieldlegal.com/law-firm-ai-policy/).

## Learning to use a general-purpose AI tool is usually overhead
A second line Opinion 512 draws is about who pays for the learning curve. The opinion indicates that a lawyer generally may not bill a specific client for the time spent learning to use a general-purpose generative AI tool, because that is more like the overhead cost of running a practice than work performed for that client [[1]](#ref-1). The time you spend figuring out how a general tool works, experimenting with prompts, or getting comfortable with its output is ordinarily a cost of doing business, not a client charge.
This tracks how the profession already treats other general competencies. A lawyer does not bill one client for the hours spent learning to use a new word processor or a general legal research platform, because that skill benefits the whole practice rather than the single matter. Opinion 512 places general-purpose AI in the same category, so the ramp-up time a firm spends becoming competent with a broadly applicable tool is generally overhead the firm absorbs [[1]](#ref-1).
There is a narrower situation the opinion treats differently, and it is worth stating carefully rather than overreading. Where a lawyer has to develop tool-specific expertise for a particular client's matter, some of that time may be billable in a way general learning is not, because it is work for that client rather than general competence. The line between general overhead and matter-specific work is fact-specific, and this article does not try to draw it for any given engagement. That is a determination for your own counsel under your jurisdiction's rules.
General learning-curve time for a broadly applicable AI tool is ordinarily overhead, not a client charge. Whether matter-specific tool work can be billed is fact-specific and best assessed with your own counsel.

## When AI cuts the time a task takes, the reasonableness cap still binds
On hourly billing, AI efficiency generally flows to the client as a lower bill. Firms that want to capture the gain look at fee structures other than pure hourly, not at billing hours that were not spent.
The situation that trips firms up most is efficiency on hourly work. When AI dramatically reduces the time a task takes, Opinion 512 treats billing the client as though the pre-AI time was spent as improper, because it charges for time not actually spent and can push the fee past what is reasonable [[1]](#ref-1). The efficiency gain, on hourly billing, generally flows to the client in the form of a lower bill, not to the firm in the form of preserved hours.
That can feel like a penalty for adopting better tools, and it is a fair thing to name plainly. A firm that invests in AI and gets faster does not, on pure hourly billing, capture that gain as higher hourly revenue for the same task. What the reasonableness standard protects is the client's interest in not paying for phantom time. Firms that want to capture efficiency gains tend to look at fee structures other than pure hourly billing, which is the subject of the next section, rather than at inflating hours on an hourly bill.
None of this makes AI-assisted work unbillable or unprofitable. It means the profit shows up as more matters handled in the same hours, or as fee arrangements that price outcomes rather than time, not as billing old hours for new speed. The reasonableness cap in Rule 1.5 is the constant, and it binds whether or not a tool made the underlying work faster [[2]](#ref-2).

## Flat and contingent fees still have to be reasonable
Not every fee is tied to hours, and the analysis shifts for flat and contingent arrangements, but it does not disappear. Under Model Rule 1.5, a flat or contingent fee must still be reasonable even if AI reduced the effort required to earn it [[1]](#ref-1) [[2]](#ref-2). The reasonableness of a fixed fee is judged against the work, the complexity, the result, and the other factors in Rule 1.5, and a tool that made the work faster is one input to that judgment rather than an automatic reason to cut the fee.
This is where some state guidance has taken a more permissive view, and it is worth attributing carefully rather than overstating. Guidance in at least one state has indicated that it is not per se unreasonable to charge the same non-hourly fee for work that AI helped produce, on the reasoning that a flat fee prices the value or outcome rather than the hours [[3]](#ref-3) [[4]](#ref-4). That is not a universal holding, and it does not mean any flat fee is safe regardless of size. It means the reasonableness of a non-hourly fee is not automatically defeated by the fact that AI made the work faster.
The practical caution is that a fixed fee can still become unreasonable if AI collapses the effort so far that the fee no longer bears a sensible relationship to the work. Rule 1.5's reasonableness requirement is the constant across fee types [[2]](#ref-2). What changes between hourly and flat billing is how the standard is applied, not whether it applies. Because states differ on how they treat non-hourly fees for AI-assisted work, the specific answer depends on your jurisdiction and belongs with your own counsel.
A flat or contingent fee must still be reasonable under Rule 1.5. Some state guidance treats charging the same non-hourly fee for AI-assisted work as not per se unreasonable, but reasonableness remains fact-specific.

## Billable versus not-billable AI scenarios at a glance
Because the analysis is fact-specific, it helps to see the common patterns side by side rather than in the abstract. The table below sorts recurring AI billing situations into what Opinion 512's application of Rule 1.5 generally supports and what it generally does not [[1]](#ref-1) [[2]](#ref-2). It is a working reference for structuring the conversation with your own counsel, not a scoring system, and it does not resolve any particular fee arrangement in any particular jurisdiction.
Read the two columns as tendencies rather than verdicts. A scenario can sit near the line, and the fee type, the jurisdiction, and the engagement terms all move the analysis. The point of the table is to make the pattern visible: real time and reasonable fees on one side, phantom time and general overhead on the other.
Scenario Generally billable Generally not billable
Time reviewing and verifying AI output Yes, it is time actually spent on the matter
Hours the task took before AI, when AI was faster No, that time was not actually spent
Learning a general-purpose AI tool No, that is ordinarily firm overhead
Directing the tool and correcting its work Yes, that is real lawyer time on the matter
A flat fee that remains reasonable for the result Yes, if reasonable under Rule 1.5 for the work
A flat fee so large it no longer fits the work No, an unreasonable fee is improper under Rule 1.5

These are general tendencies under Opinion 512 and Rule 1.5, not determinations. A specific fee depends on the arrangement, the jurisdiction, and the facts, and should be assessed with your own counsel.

## Where state guidance varies on AI and billing
ABA formal opinions are influential but not binding. States adopt, adapt, or decline them, and several state bars have issued their own guidance on lawyers using generative AI, including on how AI affects fees and costs. That means the answer to "can I bill this" can differ by jurisdiction, and checking your own state's guidance is not optional when a fee arrangement is at stake.
The Florida Bar has issued an ethics opinion addressing lawyers' use of generative AI, including how AI's effect on costs should be disclosed to clients where relevant [[3]](#ref-3). That emphasis on cost disclosure is a different foreground from a pure reasonableness framing, and it shows how a state can add expectations the ABA opinion treats more briefly. Other state guidance has taken up the non-hourly fee question, with at least one state indicating that charging the same flat fee for AI-assisted work is not per se unreasonable [[4]](#ref-4). This article does not attempt to catalog every state or to assert holdings beyond what these sources say.
The safe posture is to treat Opinion 512 and Rule 1.5 as the baseline, then read them together with your jurisdiction's own guidance, which may add requirements around disclosure or may address non-hourly fees specifically. A survey of state approaches can help you locate where your jurisdiction sits [[4]](#ref-4), but the controlling source is always the version of Rule 1.5 your licensing authority has actually adopted, interpreted by your own counsel.

- ABA Formal Opinion 512 applies Model Rule 1.5's reasonableness duty to AI-assisted billing, as influential guidance rather than a binding rule [[1]](#ref-1) [[2]](#ref-2).
- Florida's ethics opinion addresses AI's impact on costs and how they are disclosed to clients [[3]](#ref-3).
- Some state guidance indicates charging the same non-hourly fee for AI-assisted work is not per se unreasonable [[4]](#ref-4).
- Check your own jurisdiction's adopted version of Rule 1.5 and any AI-specific guidance, because requirements differ by state.

## How to document AI-assisted work so a bill is defensible
Whatever a firm decides on a given matter, the difference between a defensible bill and a disputed one is usually the record. A firm can bill honestly, capture only real time, and keep the fee reasonable, and still, months later, be unable to show which time was AI-assisted and which was full attorney work if it did not record the distinction as it happened. The steps below make that record routine, and they map to the reasonableness questions Opinion 512 turns on.
Work the sequence as part of ordinary timekeeping rather than as a separate compliance chore. The goal is not to generate paperwork. It is to be able to show, at the moment a client or a reviewer asks, that the time billed was actually spent and that the fee reflects the work. This is a structuring aid for defensible billing, not legal advice on any specific bill.

- **Record what was actually done** Capture the real time spent directing, reviewing, and correcting AI output on the matter, rather than a legacy estimate of what the task used to take [[1]](#ref-1).
- **Distinguish AI-assisted from attorney time** Note in the time entry where a tool produced a draft and where the lawyer performed independent work, so the two are visible rather than blended.
- **Keep the fee basis clear in the engagement letter** State whether billing is hourly, flat, or contingent, and address AI-related costs where relevant, so the client understood the basis for the charge [[1]](#ref-1).
- **Separate overhead from client work** Keep general tool-learning time out of client bills, and record it as firm overhead, so it never lands on a matter it should not [[1]](#ref-1).
- **Preserve the verification record** Retain evidence that AI output was checked, because that verification time is both billable and part of showing the work was competent.
- **Make the record verifiable, not just recalled** Keep the documentation in a form you can produce later, so the bill rests on a record rather than on recollection when it is questioned.
This is a documentation aid, not legal advice on any specific bill. Apply it with your own counsel and your jurisdiction's version of Rule 1.5.

## The engagement letter and being clear about the fee basis
One document does more to keep an AI-assisted bill defensible than any timekeeping habit, which is the engagement letter. The fee basis a client agreed to, hourly, flat, or contingent, frames every later question about whether a charge was proper, and Opinion 512's communication thread means that where AI affects cost, the client may need to understand that [[1]](#ref-1). A clear fee arrangement at the outset is what a firm reasons from when a bill is later questioned.
That clarity matters more, not less, as AI changes how work gets done. If a firm bills hourly, the engagement letter and the bills should reflect that AI efficiency reduces the hours, not that the firm preserves them. If a firm bills a flat fee, the letter should make the fixed-fee basis explicit, so the client understood they were paying for the outcome rather than for a count of hours, which is exactly the framing that supports a reasonable non-hourly fee even when AI made the work faster [[2]](#ref-2). The engagement letter is where the fee logic is set, and where a dispute is most cheaply avoided.
The practical implication is that AI billing is better handled as a considered part of intake than as an improvisation at invoice time. A firm that sets the fee basis clearly, addresses AI-related costs where they matter, and bills consistently with that basis is running the reasonableness analysis by design. For how AI use interacts with the broader duty to clients, see [our overview of client confidentiality and AI](https://rankshieldlegal.com/client-confidentiality-ai/).
The engagement letter sets the fee basis a bill is judged against. Where AI affects cost, Opinion 512's communication duty may mean the client needs to understand that at the outset.

## Attesting the record without overclaiming what it proves
There is a gap between billing correctly and being able to show it later, and that gap is the narrow thing verification tooling can help with. A firm can capture real time, separate overhead, and keep the fee reasonable, and still find that months on, "we only billed the review time" has become an assertion rather than a record. Timekeeping fades into recollection, and the distinction between AI-assisted and full attorney work blurs unless it was recorded as it happened.
This is where it is worth being precise about what a tool can and cannot do. RankShield Legal's approach can help produce a verifiable record of the AI-assisted steps in a matter, binding together which tool was used, when, and against the firm's own policy, so the documentation behind a bill can be checked rather than merely recalled. That supports defensible billing by making the underlying record durable. It does not decide whether a given fee was reasonable, because reasonableness under Rule 1.5 is a judgment a bar or a court makes on the facts, not a control any vendor can enforce [[2]](#ref-2).
The honest boundary is the same one this article has held throughout. A verifiable record can show what was actually done and when, which is exactly what a questioned bill needs. It cannot convert an unreasonable fee into a reasonable one, and it cannot bill for time that was not spent. What it changes is not the reasonableness standard. It changes whether a firm can show its work at the moment a client, a reviewer, or a regulator asks.
A verifiable record evidences what was done and when. It does not determine whether a fee was reasonable, which remains a fact-specific judgment under Rule 1.5. Be cautious of any tool that claims to guarantee a bill is proper.

Test yourself
## Test your AI billing knowledge
A few questions on what Opinion 512 and Rule 1.5 actually allow.

- 1 AI cuts a research task from two hours to twenty minutes. On hourly billing, what can you charge? The full two hours the task used to take The twenty minutes plus real review and verification time Nothing, because AI did the work **Answer:** The twenty minutes plus real review and verification time Opinion 512 says you may not bill time not actually spent. The bill is the real time spent, and the output being valuable does not turn unspent time into billable time.
- 2 Time spent learning a general-purpose AI tool is generally treated as what? A billable client charge Firm overhead Always prohibited to record **Answer:** Firm overhead Opinion 512 indicates a lawyer generally may not bill a specific client for learning a general-purpose AI tool, because that ramp-up is more like the overhead of running a practice than work for that client.
- 3 Under Rule 1.5, a flat fee for work AI helped produce must still be what? Automatically reduced Reasonable for the work Converted to hourly **Answer:** Reasonable for the work A flat or contingent fee must still be reasonable under Rule 1.5. Some state guidance treats charging the same non-hourly fee as not per se unreasonable, but reasonableness remains the constant.
- 4 Can software prove your AI-assisted bill was reasonable? Yes, a verifiable record settles it No, reasonableness is a fact-specific judgment a bar or court makes Only for hourly bills **Answer:** No, reasonableness is a fact-specific judgment a bar or court makes A verifiable record can show what was done and when, which supports a defensible bill, but reasonableness under Rule 1.5 is a judgment a bar or court makes on the facts, not something a vendor can enforce.
Honest self-check. There is no sign-up, and nothing is stored.

Questions answered
## Straight answers to the common questions
The questions readers ask about this topic, answered directly. **No forms, no sales pitch.**

JAMIE KLONCZ · SEO AGENCY NAPLES ************** ONLINE
Pick a question on the left, or search above. You will get the direct answer, the way an answer engine would give it.

← PREV NEXT → [REQUEST ACCESS →](https://rankshieldlegal.com/contact/)

- **Can I bill a client for work that AI helped produce?** Yes, within limits. ABA Formal Opinion 512 applies Model Rule 1.5's reasonable-fees duty to AI-assisted work, so you can bill for the time you actually spent directing a tool, reviewing its output, correcting it, and doing the parts of the work AI did not do [1][2]. What you generally cannot do on hourly billing is charge for the hours the task would have taken before AI, because that is time not actually spent. The output being valuable does not turn unspent time into billable time. The overall fee also has to stay reasonable under Rule 1.5, whatever fee structure you use. The practical discipline is to bill real time and keep the total reasonable, and to consult your jurisdiction's adopted rules, because states apply this guidance differently.
- **Can I bill the same hours if AI made the task faster?** Generally no, not on hourly billing. Opinion 512 treats billing a client as though the pre-AI process was used, when it was not, as improper, because it charges for time that was not actually spent and can push the fee past reasonable [1]. If a task that took two hours now takes twenty minutes, the hourly bill is the twenty minutes plus whatever verification and revision you genuinely performed. On pure hourly billing, the efficiency gain generally flows to the client as a lower bill rather than to the firm as preserved hours. Firms that want to capture efficiency gains tend to look at flat or other non-hourly fee structures instead of inflating hours, and reasonableness under Rule 1.5 still applies to those arrangements [2].
- **Can I bill for the time I spent learning an AI tool?** Usually not, if it is a general-purpose tool. Opinion 512 indicates that a lawyer generally may not bill a specific client for the time spent learning to use a general-purpose generative AI tool, because that ramp-up is more like the overhead of running a practice than work performed for that client [1]. It works like learning a new research platform or word processor, which benefits the whole practice rather than one matter. There is a narrower situation where developing tool-specific expertise for a particular client's matter may be treated differently, but the line between general overhead and matter-specific work is fact-specific. Keep general learning time out of client bills as overhead, and take any closer call to your own counsel under your jurisdiction's rules.
- **Does the answer change for flat or contingent fees?** The analysis shifts but the reasonableness standard does not go away. Under Model Rule 1.5, a flat or contingent fee must still be reasonable even if AI reduced the effort to earn it [1][2]. Because a non-hourly fee prices the outcome or value rather than the hours, some state guidance has indicated that charging the same flat fee for AI-assisted work is not per se unreasonable [3][4]. That is not a universal holding, and a fixed fee can still become unreasonable if AI collapses the effort so far that the fee no longer fits the work. What changes between hourly and flat billing is how the reasonableness standard is applied, not whether it applies. Because states differ on non-hourly fees, the specific answer depends on your jurisdiction and belongs with your own counsel.
- **Do I have to tell my client that AI affected my bill?** Sometimes, depending on the facts and the jurisdiction. Opinion 512 connects fee reasonableness to the communication duty under Model Rule 1.4, so where AI affects the cost of the representation, that can be something a lawyer needs to explain to the client [1]. Some states put particular weight on this: the Florida Bar's ethics opinion addresses how AI's effect on costs is disclosed to clients [3]. The clearest way to handle it is in the engagement letter, by setting out whether billing is hourly, flat, or contingent and addressing AI-related costs where they matter, so the client understood the fee basis at the outset. Treat Opinion 512 as the baseline and read it with your jurisdiction's own guidance, because disclosure expectations vary by state.
- **Can software prove that my AI-assisted bill was proper?** No, and be cautious of any tool that claims otherwise. Software can produce a verifiable record of the AI-assisted steps in a matter, binding which tool was used, when, and against the firm's policy, which supports a defensible bill by making the underlying documentation checkable rather than merely recalled. RankShield Legal's approach is aimed at that record. What it cannot do is decide whether a fee was reasonable, because reasonableness under Model Rule 1.5 is a fact-specific judgment a bar or a court makes, not a control a vendor can enforce [2]. A record cannot convert an unreasonable fee into a reasonable one, and it cannot bill for time that was not spent. What honest tooling changes is whether a firm can show what it actually did when a bill is questioned, rather than rely on recollection.

## References

- ABA Standing Committee on Ethics and Professional Responsibility. Formal Opinion 512: Generative Artificial Intelligence Tools. July 29, 2024. [https://www.americanbar.org/news/abanews/aba-news-archives/2024/07/aba-issues-first-ethics-guidance-ai-tools/](https://www.americanbar.org/news/abanews/aba-news-archives/2024/07/aba-issues-first-ethics-guidance-ai-tools/)
- American Bar Association. Model Rule 1.5: Fees. 2024. [https://www.americanbar.org/groups/professional_responsibility/publications/model_rules_of_professional_conduct/rule_1_5_fees/](https://www.americanbar.org/groups/professional_responsibility/publications/model_rules_of_professional_conduct/rule_1_5_fees/)
- The Florida Bar. Ethics Opinion 24-1: Lawyers' Use of Generative Artificial Intelligence. January 19, 2024. [https://www.floridabar.org/etopinions/opinion-24-1/](https://www.floridabar.org/etopinions/opinion-24-1/)
- Justia. State-by-State Survey of Ethics Guidance on Generative AI in Legal Practice. 2024. [https://www.justia.com/lawyers/legal-ethics/](https://www.justia.com/lawyers/legal-ethics/)

Written by
## [Jamie Kloncz](https://rankshieldlegal.com/about/)
Founder, RankShield
Jamie Kloncz is the founder of RankShield, the verifiable AI and quantum security platform behind RankShield Legal. An engineer by training, he built RankShield after his own devices and business were attacked, including an AI voice-cloning scam that targeted his family, on one conviction: unverifiable security is the real danger, so every consequential action should leave a receipt anyone can independently check.
[More about Jamie →](https://rankshieldlegal.com/about/)

Try it · Free
## Check a citation against live case-law
Paste a citation from an AI-drafted brief and see whether the case actually exists, resolved against live case-law. Free, no sign-up. Then request early access to certify a full filing.
[Try the citation checker](https://rankshieldlegal.com/ai-legal-citation-checker/)

Keep reading
## Related guides
[Legal AI When Your AI Stops Drafting and Starts Acting: The Agentic AI Governance Gap in Law Firms Read guide →](https://rankshieldlegal.com/blog/agentic-ai-governance-gap-law-firms/)[Legal AI California Is Turning AI Ethics Guidance Into Disciplinary Rules Read guide →](https://rankshieldlegal.com/blog/california-ai-disciplinary-rules/)[Legal AI What Court AI-Certification Orders Require You to Attest Read guide →](https://rankshieldlegal.com/blog/court-ai-certification-orders/)