Legal Tech and AI

AI Adoption in Law Firms: Where Small Firms Stand

AI adoption in law firms by the numbers: what Thomson Reuters, Census and Federal Reserve data show for small firms, and a 90-day plan to catch up.

AssistCall research teamPublished Reviewed by Faruque Amin, VP of Sales and Marketing6 min read

On this page
  1. What the surveys measure, and what they do not
  2. The headline numbers
  3. The firm-size gap
  4. What lawyers use it for
  5. What holds small firms back
  6. What clients know and expect
  7. What this means for a small firm
  8. Where small firms have an advantage
  9. Questions to answer before adopting anything
  10. A 90-day plan
  11. Example: a four-lawyer family law firm
  12. How to read next year's numbers
  13. Frequently asked questions
  14. References
Legal Tech and AI: AI Adoption in Law Firms: Where Small Firms StandArticle coverLEGAL TECH AND AIAI Adoption in LawFirms: Where SmallFirms StandAva for LawyersInsightsLegal Tech and AI: AI Adoption in Law Firms: Where Small Firms StandArticle coverLEGAL TECH AND AIAI Adoption in LawFirms: Where SmallFirms StandAva for Lawyers

Key takeaways

  • Generative AI use among legal professionals rose from 14% to 26% between the 2024 and 2025 Thomson Reuters reports[1].
  • Small businesses lag: fewer than 20% of US firms with four or fewer employees used AI in spring 2026, against 37% of firms with 250 or more[2].
  • Training and measurement are the gaps: 64% of professionals had no training, and only 20% of legal professionals measure ROI[3][4].
  • A small firm can close the gap with one defined use, a written policy and a 90-day review.

What the surveys measure, and what they do not

AI adoption figures are quoted constantly, and they often measure different things. Some surveys ask individuals whether they personally use generative AI. Others ask whether an organization has adopted it. Some count any AI, including spam filters and document review tools, while others mean only generative AI. Some are global, others US-only. Before comparing a headline number with your own firm, check who was asked and what counted as use.

The figures below come from three kinds of source: Thomson Reuters Institute surveys of legal and professional services, which are global; the US Census Bureau's Business Trends and Outlook Survey, which covers all US businesses; and Federal Reserve analysis of Census and other data. Read together, they give a reasonable picture of where small law firms stand.

The headline numbers

26%

of legal professionals said they were already using generative AI in 2025, up from 14% a year earlier (global survey)[1]

28%

organization-level generative AI adoption for law firms, the strongest of all professions surveyed[5]

19.8%

national AI use rate among US businesses as of May 3, 2026[6]

Legal professionals using generative AIThomson Reuters Institute, global survey
Legal professionals using generative AIComparison. 2024 report: 14%; 2025 report: 26%.14%2024 report2024 report: 14%26%2025 report2025 report: 26%Legal professionals using generative AIComparison. 2024 report: 14%; 2025 report: 26%.14%2024 report2024 report: 14%26%2025 report2025 report: 26%

Source: Thomson Reuters Institute, 2025[1]

Show the data
ItemValue
2024 report14%
2025 report26%

Legal is not a laggard overall. Thomson Reuters found law firms had the strongest organization-level adoption of the professions it surveyed[5], and a Federal Reserve analysis puts AI adoption in professional, scientific and technical services, the sector that includes law firms, at about 33%, among the highest of any sector[7].

The firm-size gap

Size is where small firms fall behind. Census data show that fewer than 20% of US firms with four or fewer employees reported using AI in spring 2026, compared with 37% of firms with at least 250 employees[2]. That matters for law because the profession is dominated by small offices: 71.8% of US law offices had fewer than five employees in 2023[8].

The gap has practical causes. Large firms have IT staff to vet vendors, knowledge management teams to build approved workflows, and budgets for enterprise contracts with stronger confidentiality terms. A solo or small firm has to do the same diligence with fewer hands, which makes a narrow, well-chosen first use more important than broad experimentation.

Figures from the cited sources; definitions of "use" differ between surveys.
SourceWho was askedFigureWhat it tells a small firm
Thomson Reuters Institute 2025Legal professionals, global26% using generative AIPeers are moving; expect clients and opponents to use it
Thomson Reuters Institute 2025Organizations, by profession28% of law firmsFirm-level adoption is still a minority
Census Bureau, spring 2026US businesses by sizeUnder 20% of firms with 4 or fewer staff; 37% at 250+Small businesses in every industry lag
Federal Reserve, 2026Sectors, from Census dataAbout 33% in professional servicesLaw sits in one of the highest-adopting sectors

What lawyers use it for

Among legal professionals who use generative AI, the top uses are document review (74%), legal research (73%) and document summarization (72%)[9]. These are internal tasks where a lawyer checks the output before anything reaches a client or court. Client-facing uses, such as intake or chat, are less common and carry more professional responsibility risk, which is why state bars have begun to address them directly.

What holds small firms back

  • Training: although 89% of professionals see use cases for generative AI, 64% have received no training on using it for their work[3].
  • Measurement: only 20% of legal professionals surveyed are measuring the return on investment of generative AI[4], so many firms cannot tell whether a tool is worth keeping.
  • Confidentiality: lawyers must understand where client information goes before using a tool, which takes time to check.
  • Accuracy: hallucinated citations have led to sanctions, and the fear is reasonable.
  • Time: the people who would evaluate tools are the same people billing hours.

What clients know and expect

Clients are often unaware either way. Thomson Reuters found 71% of law firm clients did not know whether their firms used generative AI tools, while 59% of corporate legal clients wanted their outside firms to use it[10]. For consumer practices, the signal is weaker, but the direction is similar: clients expect faster responses and clearer updates, and they are unlikely to ask how you produced them. That puts the burden on the firm to decide what to tell clients and when consent is needed.

What this means for a small firm

Being behind large firms is not a crisis, and rushing in carries its own risk. The useful conclusion is narrower: peers are adopting for internal drafting and review work, clients will not stop you, and the firms that benefit are those that pick a defined use, train people and measure the result. A two-lawyer firm does not need an AI strategy. It needs one well-chosen workflow, a short written policy and a date to review it.

Where small firms have an advantage

Small firms are slower to adopt, but they can move faster once they decide. There is no committee, no procurement cycle and no need to roll a tool out to hundreds of people. A solo lawyer can test a tool on a closed matter this week, write a one-page policy next week and know within two months whether it saves time. Small firms also tend to have simpler data: fewer systems, fewer integrations and a clearer picture of where client information lives, which makes the confidentiality review easier than at a large firm with decades of legacy systems.

Questions to answer before adopting anything

  • Which task will this tool do, and how will we know if it did it well?
  • What client information, if any, will go into it, and what do the terms say about retention and training?
  • Who checks the output, and what exactly do they check?
  • How will we bill for the time it saves?
  • What will we tell clients, and do we need consent for this use?
  • When will we review the decision, and what result would make us stop?

Write the answers down before the first use. They become the core of the firm's policy and give the 90-day review something concrete to measure against.

A 90-day plan

  1. Days 1 to 15: pick one task

    Choose a repetitive internal task with easy verification, such as summarizing deposition transcripts you already have or drafting routine letters from your own templates.

  2. Days 1 to 15: check the terms

    Read the tool's terms of use and privacy policy. Confirm whether inputs are retained or used for training and whether that fits your confidentiality duties.

  3. Days 16 to 30: write a short policy

    One page: approved tool, approved task, what data may be entered, who checks output and how time is billed.

  4. Days 31 to 75: run it and log it

    Log each use: task, time taken, corrections needed. Keep the log simple enough that people actually fill it in.

  5. Days 76 to 90: decide

    Compare time and error rates with the old method. Keep, adjust or drop the tool, and choose whether to add a second task.

Example: a four-lawyer family law firm

Example (fictional): a four-lawyer family law firm tries generative AI for one task: first-draft summaries of financial disclosure documents already in the file. The office manager checks the vendor's terms and the firm limits use to an enterprise account that does not train on inputs. Over 60 days, the log shows summaries take about a third of the previous time, with one in five needing a material correction, mostly misread dates. The firm keeps the tool, adds a date-check step to its policy and decides not to expand to client-facing uses for now.

  • Check whether a figure counts individuals or organizations.
  • Note whether the survey is global or US-only, and whether it includes small firms.
  • Look for the definition of use: any AI, or generative AI only.
  • Prefer government statistics and established research publishers over vendor surveys of their own customers.
  • Compare like with like across years; changes in survey method can look like adoption jumps.

Frequently asked questions

What percentage of law firms use AI?

Thomson Reuters reported organization-level generative AI adoption of 28% for law firms in 2025, in a global survey[5]. Figures differ by survey and definition.

Are small law firms behind on AI?

Small businesses generally are: fewer than 20% of US firms with four or fewer employees used AI in spring 2026, against 37% of the largest firms[2].

What do lawyers mainly use generative AI for?

Document review, legal research and document summarization are the top uses among legal professionals who use it[9].

Should a small firm start with client-facing AI?

Usually not. Internal tasks with easy verification are a safer first step; client-facing uses need stronger safeguards and clear disclosure.

References

  1. Thomson Reuters Institute (2025). 2025 GenAI report: Executive summary for legal professionals (opens in a new tab) โ†‘

  2. U.S. Census Bureau (2026). Large Firms With at Least 20 Employees Biggest AI Users (opens in a new tab) โ†‘

  3. Thomson Reuters (2025). Generative AI Adoption Nearly Doubles as Professional Services Reach Crossroads (opens in a new tab) โ†‘

  4. Thomson Reuters Institute (2025). 2025 GenAI report: Executive summary for legal professionals (opens in a new tab) โ†‘

  5. Thomson Reuters Institute (2025). 2025 GenAI report: Executive summary for legal professionals (opens in a new tab) โ†‘

  6. U.S. Census Bureau (2026). Large Firms With at Least 20 Employees Biggest AI Users (opens in a new tab) โ†‘

  7. Board of Governors of the Federal Reserve System (2026). Monitoring AI Adoption in the U.S. Economy (FEDS Notes) (opens in a new tab) โ†‘

  8. U.S. Census Bureau (2023). County Business Patterns 2023: U.S. detail file (cbp23us) (opens in a new tab) โ†‘

  9. Thomson Reuters Institute (2025). 2025 GenAI report: Executive summary for legal professionals (opens in a new tab) โ†‘

  10. Thomson Reuters Institute (2025). 2025 GenAI report: Executive summary for legal professionals (opens in a new tab) โ†‘

Book a demo

See Ava handle your calls.

Book a short demo. Weโ€™ll walk through your calls and workflow, and agree what Ava should take care of.

Book a demo