Demo Spellbook, get an Oura Ring 5*

See how Spellbook helps legal teams draft contracts faster than ever before.

Book Demo
Close modal

Generative AI's Impact on Legal Practice

Last updated: Nov 10, 2026
Written by
Kurt Dunphy
Kurt Dunphy
Generative AI's Impact on Legal Practice

Generative AI for legal practice now does real work in five places: contract drafting, contract review, due diligence, legal research and client correspondence. In each of them it compresses the mechanical layer of legal work while judgment, supervision and verification stay with the lawyer. Courts have now sanctioned filings built on generative AI output that nobody checked, which is why every workable use of generative AI for legal work pairs a drafting gain with a named verification step.

Teams that get value from generative AI have settled three things before the first matter: which workflow the technology owns, who verifies its output, and what happens when it is wrong. Teams that have left those undecided carry the real exposure, because individual lawyers then paste client language into consumer chatbots with no policy behind it. Commercial teams at law firms close that gap by scoping one workflow at a time rather than by buying a platform and hoping adoption follows.

[cta-1]

What Generative AI Is in Legal Practice

Generative AI in legal practice is software that produces new text, such as a clause, a summary or a memo, from statistical patterns learned across large bodies of documents, rather than sorting documents that already exist into categories. The distinction matters because the two technologies fail differently: a classifier mislabels a clause it has seen, and a generative model writes a confident sentence about a case that does not exist.

Model Generations and What to Check

The models behind these tools have moved through several generations since 2022, and the current generation is materially better at long documents than the one most firms first tested. Where a vendor does not publish which model version sits behind a feature, evaluate the capability claim rather than the model name.

  • Ask which model version sits behind the feature, and ask again at renewal, because the answer changes within a contract term.
  • Test the capability on your own long documents, since the gains between model generations show up most on long agreements.

Generative AI Compared With Traditional Legal AI

Traditional legal AI classifies and extracts. It reads a document set, assigns each clause to a category, and pulls named fields into a table. The output is a label or a value, and it is traceable to a span of text in the source document.

Generative AI writes. Given a prompt and a precedent, it produces language that did not exist before. An extraction error is visible against the document, and a generation error reads like competent drafting until someone checks the authority behind it.

  • Traditional legal AI answers which category a clause belongs to, and the answer points back to text in the document.
  • Generative AI answers what the clause should say, and the answer has no source span to point at.
  • The verification differs accordingly: an extraction is checked against the document, a generation against the authority it asserts.

How Generative AI Is Changing the Legal Industry

Generative AI in the legal industry has changed how first drafts and first reviews get produced, it has not changed who is accountable for them, and it has added a verification step that did not exist in 2022. Adoption rose fastest in document-heavy transactional work, where the input is a contract and the output is a redline a lawyer reads line by line.

What Changed at the Workflow Level

The starting point moved. A first-pass nondisclosure agreement (NDA) or a summary of a full data room now arrives before a lawyer has read the first page, and the lawyer’s time moves from production to checking and positioning.

  • The first draft stopped being the bottleneck on standard commercial paper, which moved the bottleneck to review capacity.
  • Coverage rose on document sets that were previously sampled, because reading all of them stopped being a timeline question.
  • A verification step entered workflows that did not previously have one, and it has to be staffed rather than assumed.

What Did Not Change

Accountability for the output stays exactly where it was. The signed agreement still reflects a decision a named lawyer made.

  • The duty of competence applies to the finished work product regardless of what produced the first version of it.
  • Supervision obligations extend to output a junior lawyer accepted from a tool, the same way they extend to a junior lawyer’s own drafting.

The Governance Gap

The governance layer is where the industry is furthest behind its own usage. Written AI use policies, training, matter-level disclosure practices and any measurement of output quality all lag the number of lawyers already using these tools. The ABA Task Force on Law and Artificial Intelligence has treated that lag as a practice-management problem rather than a technology problem. As a result, the risk today usually sits in undocumented individual use, where lawyers already apply generative AI to their own day-to-day work without any written policy behind it, rather than in sanctioned deployments that a firm actually tracks.

Generative AI Use Cases in Commercial Legal Work

Use case What generative AI produces What the lawyer still does
Contract drafting A first-pass clause or agreement from a prompt and precedent Sets the position, checks the fallback, approves the language
Contract review A list of flagged risks and proposed redlines Decides which risks matter on this deal
Due diligence Summaries and extracted terms across a document set Confirms the extraction against the source document
Legal research Candidate authorities and a summary of them Reads and verifies every citation
Correspondence A plain-language explanation of a term or issue Confirms accuracy and sends it

Contract Drafting and First-Pass Clause Generation

Drafting has structured inputs: a playbook position and a precedent, with an output that is checkable against a precedent the team already trusts. A model given those inputs produces a serviceable first draft of a standard commercial agreement, and the work that remains is the work that was always the lawyer’s: choosing the position and deciding how far the fallback goes.

A fluent draft can still be internally inconsistent, usually a defined term used in two senses or a notice period that contradicts the term in another clause, and catching that depends on how well generative AI for legal contracts is applied to the clause library and playbook behind the draft.

  • Read the defined terms and the cross-references first, since that is where fluent drafts break.
  • Keep the precedent in the loop, because a draft produced without one reverts to market-average language.

Contract Review and Redlining

Review is where the volume problem bites hardest: a team facing more third-party paper than review hours triages instead of reading. Generative AI reads the counterparty draft against a position and proposes redlines in the document.

Whose paper it is determines how useful this gets. On your own template, deviations are the whole signal. On third-party paper, the tool flags every departure it finds and the lawyer decides which of them matter on this deal, which is judgment that does not transfer. Teams negotiating from the counterparty’s draft run into this split first.

  • On your own template, treat every flagged deviation as in scope for review.
  • On third-party paper, rank the flags against the deal position before reading any of them in full.

Due Diligence and Multi-Document Review

Diligence changes coverage: a model can read every agreement in a data room rather than the sample a timeline allows, extracting the change-of-control, assignment and exclusivity provisions.

Confirm the extracted term against the source document, because an extraction attached to the wrong clause reads exactly like a correct one in a summary table. Allocating that verification across a deal team is the part of multi-document review that needs designing before the first data room opens.

  • Verify every extracted provision that feeds a disclosure schedule or a closing condition.
  • Spot-check the documents the tool reported as having no relevant provision, which is where silent misses sit.

Legal Research and Summarization

Research use carries a verification burden on every output, because published evidence on these tools is not reassuring. Magesh and colleagues ran a preregistered evaluation of retrieval-augmented legal research tools and found they hallucinate between 17% and 33% of the time. At that published error rate, unverified research output cannot go into a filing.

Generative AI is useful for finding candidate authorities and unsafe for stating what they hold. The lawyer reads each case. Treating the output as a research lead rather than a research answer is the distinction that keeps AI-assisted research inside the duty of competence.

  • Pull every authority from a primary database before it reaches a draft, and confirm the holding against the text.
  • Check the currency of each authority separately, since a model with no live authority check cannot see a reversal.

Client and Counterparty Correspondence

A plain-language explanation of an indemnity or a limitation of liability, drafted for a commercial counterpart who is not a lawyer, is work a model does well and work a partner is expensive to do.

This output leaves the firm. An explanation that softens a cap or omits a carve-out becomes a representation the client relied on, so the review standard for outbound correspondence is the same standard as for the clause it describes, not a lighter one.

  • Read the explanation against the operative clause, not against the summary the tool produced of it.
  • Keep the carve-outs and the exceptions in, because the simplification a model makes is usually to drop them.

What These Use Cases Have in Common

Every one of the five pairs a production gain with a verification step, and the gain tracks how structured the input is.

  • Where the input is a document and the output is checkable against it, the gain is largest and the check is cheapest.
  • Where the output asserts something outside the document, such as an authority or a market position, the check costs more than the drafting it saves.

What Generative AI Still Cannot Do in Legal Work

The limitations of generative AI in legal work are structural rather than temporary: the technology predicts likely text, so it cannot know whether a case is still good law, cannot weigh a client’s risk appetite, and cannot take responsibility for advice. A model that has never been told which fallback the client accepted will propose the one that appears most often in its training data.

  • Currency is a retrieval problem rather than a generation problem, and a model with no live authority check cannot tell you a decision was reversed last month.
  • Risk appetite is client-specific and mostly unwritten, which is why a tool trained on market language proposes market terms to a client who has decided not to accept them.
  • Accountability is a professional obligation that does not delegate, and no model generation changes that.

Where the Human Step Sits

The human step is not a formality. It is the step where someone who knows the matter decides whether the output is right for this client, on this deal, under current law, and that decision is recorded against a name. The question about replacement therefore turns on accountability rather than on capability.

  • Place it before the output leaves the team, not after a client or a court has seen it.
  • Give it to someone with the matter context, because a check run by someone without it catches only mechanical errors.

[cta-2]

Risks and Ethical Duties of Generative AI in Legal Work

The generative AI legal ethics risks lawyers actually face are documented rather than speculative: a public database of judicial decisions addressing AI-hallucinated material held 2,266 cases on November 5, 2026, 1,535 of them in the United States. ABA Formal Opinion 512, issued on July 29, 2024, maps these risks onto six existing duties: competence, confidentiality, communication, candor toward the tribunal, supervision, and fees.

Hallucinated Citations and the Verification Step

Fabricated authority is the risk with the longest public record. The database of judicial decisions addressing hallucinated material tracks matters where a court found a party relied on content an AI tool invented, and the count has grown steadily rather than tailed off as the tools improved.

The canonical matter is Mata v. Avianca, where counsel filed a brief citing cases that did not exist and then submitted fabricated opinions when the court asked for them. The docket in that case is worth reading for what the sanctions turned on, which was not the use of the tool but the failure to check its output and the response once challenged.

What the Verification Step Involves

The verification step is concrete rather than attitudinal, and a team that writes it down is a team that can show it happened. That record is what turns a hallucination caught internally into a working process rather than a near miss, and it is the record a court asks for when a filing is challenged.

  • Pull every cited authority from a primary database and confirm that it exists.
  • Confirm it says what the draft says it says, and confirm it has not been reversed or superseded.
  • Record who ran the check and when, naming the person rather than the team.
  • Apply the same check to quotations and pin cites, since a real case can still be given a fabricated page reference.

Client Confidentiality and Where the Data Goes

ABA Model Rule 1.6 does not change because the recipient of client information is a model. Formal Opinion 512 analyzes the point directly: a lawyer who enters client information into a tool that retains it, trains on it, or exposes it to anyone outside the relationship may need informed client consent before doing so.

The practical distinction is between a consumer chatbot and a tool bound by its contract. A general-purpose consumer product may retain conversations and use them in training by default, while a legal tool can commit contractually to zero data retention, which means the inputs are not stored after the request completes. That difference is a diligence question for procurement, and the privilege analysis behind it is worth settling before the first matter rather than after.

  • Ask for the retention and training terms in writing, and read them against the engagement terms you have with the client.
  • Decide which matters require client consent before any client language reaches a tool, and record the decision in the file.

Disclosure to Courts and Clients

Disclosure runs on two tracks. A growing number of judges have issued standing orders requiring a certification about generative AI use in filings, and those orders differ by courtroom, so the obligation is local and has to be checked per matter rather than per firm.

Client disclosure is the second track. Formal Opinion 512 treats the duty of communication as engaged where the use of generative AI is material to the representation, and the duty around fees as engaged where a tool changes how long the work takes. Neither obligation resolves to a single sentence in an engagement letter, and some state-level guidance remains in draft, which is why the disclosure position is something to document and revisit rather than settle once.

  • Check the standing orders for the specific court and judge at the start of each matter, because a firm-level answer does not hold.
  • Separate the fee question from the disclosure question, since a tool that changes the hours changes the engagement terms even where no filing is involved.
  • Revisit the position when state guidance moves from draft to final, and date the version the team is working to.

Copyright and Ownership of Generative AI Output

Copyright protection in the United States requires human authorship. The Copyright Office’s Part 2 report on copyrightability concludes that material generated entirely by a machine is not protectable, while a work containing AI-generated material can be protected to the extent of the human contribution to it.

For commercial legal work the consequence is narrow. Contract language is rarely where a client’s copyright value sits, so the exposure is usually in the client’s own AI-assisted product or marketing output.

  • Advise the client to record what the human contributed, in the form of edits, selection and arrangement.
  • Treat a client who cannot describe that contribution as a client who should not be relying on registration for the work.

How to Start Using Generative AI in Legal Practice

To start using generative AI in legal practice, pick one bounded workflow, write the use policy before the first matter, choose a tool built for that workflow, name who verifies the output, and measure the change. The order matters: teams that choose the tool first end up writing a policy around whatever the tool happens to do.

“Théâtre D’opéra Spatial” by Jason Allen
“Théâtre D’opéra Spatial” by Jason Allen was created using generative AI via Midjourney. It won first place in the digital category at the Colorado State Fair.

1. Pick One Bounded Workflow

  • Choose the workflow with the highest volume and the most structured output, which in most commercial teams is first-pass review of inbound third-party agreements.
  • Define the boundary in writing: which agreement types are in scope, which value threshold is out, and which matters are excluded entirely.
  • Leave litigation filings and novel advisory work out of the first workflow, because the verification cost there exceeds the drafting gain.

2. Write the AI Use Policy Before the First Matter

  • State which tools are approved, and say plainly that client information does not go into anything outside that list.
  • Record the confidentiality basis for each approved tool, including whether inputs are retained and whether client consent is required.
  • Name the disclosure position for courts and for clients, and say who decides when a matter needs one.

3. Choose a Tool Built for the Work

The procurement question is fit rather than capability, and the shortlist is worth weighing against the way tools for commercial lawyers actually divide by workflow.

  • Match the tool to the workflow rather than to the breadth of the feature list, since a contract tool and a research tool fail in different ways and need different checks.
  • Test it on your own documents rather than on a vendor sample, and read the output against a precedent your team already trusts.
  • Ask what happens to the inputs, and treat an unclear answer as a reason to stop.

4. Name the Verification Step and Its Owner

  • Write down what verification means for this workflow, for example that every extracted term is confirmed against the source clause before it reaches a summary table.
  • Assign it to a named role rather than to the team, and make it visible in the matter file so supervision under Model Rule 5.1 is demonstrable.
  • Decide what happens when the check fails, including who is told and whether the matter pauses.

5. Measure the Time and Quality Change

  • Capture cycle time per agreement on a consistent cut, for example inbound NDAs received between two dates, measured from receipt to returned redline.
  • Capture a quality measure the team can actually collect, such as the number of departures the reviewer added that the tool missed, counted per matter.
  • Compare against a pre-adoption baseline on the same agreement type, because a comparison across different agreement types measures the mix rather than the change.

Where Spellbook Fits in Generative AI for Legal Work

Spellbook is an AI suite for transactional legal work, which means its use of generative AI is deliberately narrow: commercial contract drafting and review, done in the document rather than in a separate chat window. It works in Word and Google Docs, and the Draft and Review features are grounded in the team’s own playbooks and precedents plus 20M+ real market contracts. Spellbook drafts and auto-completes using GPT-5.

In the above image, Spellbook used GPT-3 to analyze an entire contract and generate a new clause in green after having been provided with the prompt in blue.

Where Spellbook Stops

That scope is a deliberate limit rather than a gap. Spellbook is a commercial-contract specialist, not a legal-research or litigation platform, which is why the research use case above is not a Spellbook use case. A team that needs authority checking and a team that needs faster redlines are buying two different things.

What the Features Do

Within contracts, each feature covers a defined part of the workflow.

  • The Review feature reads a counterparty draft against a position and flags the departures in the document.
  • The Draft feature produces first-pass language from a prompt and a precedent.
  • The Playbooks feature holds the team’s positions and fallbacks, so suggestions reflect what this team accepts rather than what the market average accepts.
  • The Compare to Market feature places a clause against real market language.
  • The Proofread feature catches the mechanical errors a redline introduces.
  • The Associate feature executes multi-step instructions across every document in a project at once, which is the multi-document diligence work described above.
  • The Platform brings these together with Spellbook Ask for questions against a document set.

Verification applies to this tooling as much as to any other. A suggested redline is a proposal, the lawyer decides which departures matter on the deal, and the tool can help ensure the mechanical layer is consistent without taking a position on risk appetite. Pricing is custom per seat and scales by team size, and there is a free seven-day trial.

Try Spellbook free.

[cta-3]

Frequently Asked Questions About Generative AI in Legal Practice

Is It Ethical for Lawyers to Use Generative AI?

Yes, within existing duties. ABA Formal Opinion 512 treats generative AI use as governed by the rules already in place on competence, confidentiality, communication, candor toward the tribunal, supervision, and fees, rather than by a new prohibition. The ethical question is therefore not whether to use it but whether the specific use satisfies those duties, which turns on what the tool does with client information and who verifies its output.

Do Lawyers Have to Tell Clients They Used Generative AI?

Sometimes, and it depends on materiality. The duty of communication is engaged where the use of generative AI is material to the representation, and the duty around fees is engaged where a tool changes how long the work takes. Courts add a separate obligation: a number of judges have issued standing orders requiring certification about AI use in filings, and those orders differ by courtroom, so the filing-side position has to be checked per matter.

Can Generative AI Output Be Copyrighted?

Not the machine-generated part. Copyright protection in the United States requires human authorship, so material produced entirely by a model is not protectable, while a work containing AI-generated material can be protected to the extent of the human contribution. The practical test is whether the author can describe and show what the human contributed.

What Is the Difference Between Generative AI and Legal AI Tools?

Generative AI is the underlying capability, and a legal AI tool is a product built around it for a defined workflow. The older generation of legal AI classified and extracted, producing labels and field values traceable to the source text. Generative tools write new language, which is more useful and harder to check, because a generation error reads like competent drafting until someone verifies the authority behind it.

Will Generative AI Replace Lawyers?

No, not on current capability. The capability gap is in the parts of legal work that do not delegate. A model cannot confirm a decision is still good law without a live authority check, cannot weigh a client’s risk appetite, and cannot hold professional responsibility for advice. What it changes is the composition of the work, moving time from producing a first draft toward checking, positioning and deciding, which is a change in the task mix rather than in who is accountable.

50+ Prompts for Contract Review and Drafting
The Morning Paper for Lawyers Who ♥️ Al
NEWSLETTER
A Hands-On Guide for In-House Teams

Download: Generative AI's Impact on Legal Practice

Please enter your work email address (not gmail, yahoo, etc.)
*Required
Oops! Something went wrong while submitting the form.
Close modal

Get started for free

Need to discuss your use case? Contact us.

please enter your business email (not gmail, yahoo, etc)
*Required

Submission Received

Thank you for your interest!

Oops! Something went wrong while submitting the form.

Join over 5,000 legal teams using Spellbook

please enter your business email (not gmail, yahoo, etc)
*Required
Close modal

Submission Received

Thank you for your interest!

Oops! Something went wrong while submitting the form.
Spellbook

One Moment

We're connecting you with the best rep