The Legal AI Workflow Integration Report 2026: How Law Firms and Legal Departments Are — and Are Not — Embedding AI Into Practice Management, Matter Tracking, and Billing Systems

Research briefing — a review of public sources. The Legal Stack did not conduct a survey, benchmark or interviews for this piece.

Executive summary

The question in this report's title is how deeply AI has been wired into the systems law firms and legal departments run on: practice management, matter management, document management, and time and billing. Public evidence cannot answer it precisely. We found no published survey that measures how often AI output is written back automatically into a system of record versus copied across by hand. Adoption surveys measure use, not integration.

The public record does show three things. AI use is rising quickly, especially at larger firms. The vendors that own the systems of record (Clio, NetDocuments, Intapp, Aderant) are now building AI into those systems directly, which removes some of the need to connect standalone tools. And ABA Formal Opinion 512 sets billing and confidentiality rules that any integration, especially in time entry, has to support. This briefing sets out that record and then gives clearly labelled analysis and recommendations.

What adoption data does and does not show

The most widely cited U.S. benchmark is the ABA's 2024 Legal Technology Survey Report. As reported by LawSites in March 2025, 30% of respondents said they use AI, up from 11% in 2023. At firms with 100 or more attorneys the figure was 46%, up from 16% in 2023. Among solo practitioners it was 18%. Accuracy was the top concern for 75% of respondents, up from 58% the year before.

Other surveys show higher numbers because they ask different questions. Clio's 2024 Legal Trends Report found that 79% of legal professionals used AI in some form, up from 19% in 2023. It also estimated that up to 74% of hourly billable tasks could be exposed to automation. The Thomson Reuters 2025 Generative AI in Professional Services report found that 26% of legal organisations were actively using generative AI, up from 14% in 2024. It also found that 41% of law firms had policies governing its use.

These figures differ because they measure different things: individual use, organisational deployment, and policy. None of them measures whether AI tools read from or write to practice management, matter or billing records. Our view is that the title's question, how much AI work actually lands in systems of record, has no reliable public answer. The figures in earlier versions of this briefing, including write-back rates by firm size and minutes of remediation per time entry, were not drawn from any verifiable source and have been removed.

Where AI is moving into systems of record

Vendor announcements are the best public evidence of integration. They are claims about capability, not proof of adoption, but they show where the architecture is heading.

Practice management. Clio unveiled Clio Work, an AI workspace for legal research and analysis, at ClioCon in October 2025, according to LawSites. It launched at $199 per user per month, was initially available only to Clio Manage subscribers, and combines Clio's research corpus with matter-level context from a firm's own documents, notes and contacts. In April 2026 Clio added agentic, multi-step capabilities. On 21 April 2026 it made Clio Work available as a standalone product for solo, small and mid-sized firms. It followed Clio's US$1 billion acquisition of vLex, announced in June 2025 and completed in November 2025, alongside a US$500 million Series G at a US$5 billion valuation, according to Clio's 2025 year in review.

Document management. NetDocuments launched ndMAX Studio in February 2025 with 12 AI applications, including a Playbook Generator, an Issues List Generator and Deposition Prep, according to IT Brief. Import and export tools bring the apps into its document management system.

Time and billing. Aderant announced AI features for its iTimekeep time-capture product in August 2023. One is a Passive Time Assistant that captures activity and matches it to client and matter. The other is a Time Narrative Assistant, trained on accepted and rejected bills, that flags entries needing edits for compliance. Intapp's Intapp Time release of 21 May 2026 uses AI to predict phase, task and activity codes from the narrative. It also lets administrators configure rules "covering outside counsel guidelines, restricted terms, and block billing," with AI-suggested rewrites at the point of entry, and Intapp says agentic workflows are coming. Intapp also claims the product helps drive "$12 billion in incremental billings each year." The release gives no methodology for that figure, which should be treated as marketing.

Analysis. The "clipboard workflow" problem is increasingly being solved from the system-of-record side rather than by standalone AI tools learning to write back. For smaller firms, that favours buying AI from the platform they already run. For larger firms with Elite, Aderant or Intapp back offices and an iManage or NetDocuments DMS, integration still means stitching together several vendors' AI layers. Matter numbers, billing codes and security classifications have to survive every handoff between them.

The billing constraint: what the rules actually say

The billing integration problem is partly a rules problem. ABA Formal Opinion 512, issued 29 July 2024, is the ABA's first formal opinion on lawyers' use of generative AI. An ABA Law Practice Today analysis summarises its fee guidance as follows:

A lawyer billing hourly "must charge only for actual time spent," even when AI makes the work faster.
A flat or contingent fee may be unreasonable if a lawyer can finish the work much faster with an AI tool.
Lawyers generally may not bill clients for time spent learning an AI tool, unless the client specifically asked for that tool to be used.
Direct costs of an AI tool may be passed through, but general overhead should not be, absent prior disclosure and agreement.

The opinion itself also says lawyers must disclose their AI practices if a client asks, or if disclosure is required by the engagement agreement or the client's outside counsel guidelines.

The market context makes this practical, not theoretical. The Thomson Reuters Institute and Georgetown Law's State of the US Legal Market 2026, as reported by Legal IT Insider in January 2026, found that about 90% of legal dollars still flow through standard hourly arrangements (per Thomson Reuters Legal Tracker data). It also found that most clients do not know whether or how their outside firms use generative AI. The ACC/Everlaw 2025 survey of 657 in-house professionals found that 59% had seen no generative AI savings from their outside firms yet.

Analysis. If clients cannot see AI's effect on invoices and most billing is still hourly, the time-entry system is where AI's economics either show up or disappear. Tools that enforce outside counsel guidelines at entry, like the Intapp feature above, address narrative compliance. We found no public documentation of an industry-standard field or code for flagging AI-assisted time in e-billing formats. Firms that want that visibility currently have to build it themselves through custom fields, matter-level reporting or engagement-letter terms.

Security and confidentiality as the integration gate

Write-back integration means giving an AI system read and write access to client data, so confidentiality rules shape what is possible. Opinion 512 says a client's informed consent is required before information relating to the representation is put into a self-learning generative AI tool. It also says that "merely adding general, boiler-plate provisions to engagement letters" purporting to authorise AI use "is not sufficient."

Courts have started to address the platform-terms question. In United States v. Heppner (S.D.N.Y., February 2026), Judge Jed Rakoff held that a defendant's own conversations with a consumer AI chatbot were not protected by privilege or work product. Part of his reasoning was that the consumer privacy terms allowed the provider to use inputs for training and disclose them to third parties, according to Orrick's analysis. Orrick notes that enterprise platforms with negotiated confidentiality terms may be treated differently, but this has not yet been tested.

Analysis. Integration projects and security review are the same project. The contract terms that govern training use, retention and third-party disclosure determine whether a connector into the matter system is defensible. These are the terms a firm's general counsel and its clients' outside counsel guidelines will scrutinise.

Recommendations (analysis)

Measure integration, not just use. Before buying more AI, inventory which tools read from and write to the DMS, practice or matter management system, and time entry, and which depend on manual transfer. No public benchmark will do this for you.
Prefer AI from the system of record where it is adequate. Platform-native AI (Clio Work, ndMAX, Intapp Time) inherits matter context and security models. Standalone tools need a documented integration path before deployment.
Configure billing rules at entry. Load outside counsel guidelines, restricted terms and block-billing rules into time-entry tools where your platform supports it. Treat AI-suggested narratives as drafts the timekeeper owns.
Align billing policy with Opinion 512. Write down how AI-assisted hourly work is recorded, how AI costs are charged, if at all, and what you will disclose when clients ask or their guidelines require it.
Make enterprise terms a precondition for write-back. No connector into client matter data without contractual limits on training use, retention and disclosure, given the reasoning in Heppner.

Sources

ABA Journal / LawSites (Bob Ambrogi), "ABA Tech Survey Finds Growing Adoption of AI in Legal Practice, with Efficiency Gains as Primary Driver" (7 Mar 2025): https://lawnext.com/2025/03/aba-tech-survey-finds-growing-adoption-of-ai-in-legal-practice-with-efficiency-gains-as-primary-driver.html
LawSites, "AI Adoption by Legal Professionals Jumps from 19% to 79% in One Year, Clio Study Finds" (7 Oct 2024): https://lawnext.com/2024/10/ai-adoption-by-legal-professionals-jumps-from-19-to-79-in-one-year-clio-study-finds.html
LawSites, "Thomson Reuters Survey: Over 95% of Legal Professionals Expect Gen AI to Become Central to Workflow Within Five Years" (15 Apr 2025): https://www.lawnext.com/2025/04/thomson-reuters-survey-over-95-of-legal-professionals-expect-gen-ai-to-become-central-to-workflow-within-five-years
LawSites, "Clio Work, Clio's AI Workspace, Is Now Available to Solo and Smaller Law Firms as a Standalone Product" (Apr 2026): https://www.lawnext.com/2026/04/clio-work-clios-ai-workspace-is-now-available-to-solo-and-smaller-law-firms-as-a-standalone-product.html
IT Brief UK, "NetDocuments unveils ndMAX Studio with 12 AI legal apps" (20 Feb 2025): https://itbrief.co.uk/story/netdocuments-unveils-ndmax-studio-with-12-ai-legal-apps
Aderant, "Aderant Revolutionizes Time Tracking for Law Firms with New AI-Powered Solutions for iTimekeep" (8 Aug 2023): https://www.aderant.com/news-pr/aderant-revolutionizes-time-tracking-for-law-firms-with-new-ai-powered-solutions-for-itimekeep/
Business Wire / Intapp, "New Intapp Time Release Brings AI Deeper Into the Time Entry Workflow to Capture More Billable Hours and Protect Earned Revenue" (21 May 2026): https://www.businesswire.com/news/home/20260521267843/en/new-intapp-time-release-brings-ai-deeper-into-the-time-entry-workflow-to-capture-more-billable-hours-and-protect-earned-revenue/
American Bar Association, "Formal Opinion 512: Generative Artificial Intelligence Tools" (29 Jul 2024): https://www.americanbar.org/content/dam/aba/administrative/professional_responsibility/ethics-opinions/aba-formal-opinion-512.pdf
ABA Law Practice Today, "The Reasonableness of Fees When Using AI" (Dec 2024): https://www.americanbar.org/groups/law_practice/resources/law-practice-today/2024/december-2024/the-reasonableness-of-fees-when-using-ai/
Legal IT Insider, "Thomson Reuters' State of the US Legal Market report – Record profits and increasingly unstable ground" (7 Jan 2026): https://legaltechnology.com/thomson-reuters-state-of-the-us-legal-market-report-record-profits-and-increasingly-unstable-ground/
Business Wire / ACC and Everlaw, "64% of In-House Counsel Expect GenAI to Reduce Reliance on Outside Counsel; 50% Expect Lower Outside Counsel Costs" (14 Oct 2025): https://www.businesswire.com/news/home/20251014382116/en/64-of-In-House-Counsel-Expect-GenAI-to-Reduce-Reliance-on-Outside-Counsel-50-Expect-Lower-Outside-Counsel-Costs-ACCEverlaw-Survey-Finds
Orrick, "Court Rules AI Conversations Are Not Privileged: What United States v. Heppner Means for You" (Mar 2026): https://orrick.com/en/Insights/2026/03/Court-Rules-AI-Conversations-Are-Not-Privileged-What-United-States-v-Heppner-Means-for-You
Clio, "Clio 2025 Year in Review" (Dec 2025): https://www.clio.com/blog/clio-2025-year-in-review/