What the 2026 ILTA survey and Morgan & Morgan’s transformation chief tell personal injury firms about where AI belongs, and where it does not
Ask a room of lawyers to name the top AI product for law firms and you will hear the same three or four names. Purpose-built legal AI platforms fill the trade press, the vendor emails and the CLE agendas. So, which one do firms use most?
None of them. The International Legal Technology Association recently released its 2026 Technology Survey. The product most firms use or explore is Microsoft 365 Copilot, a general office assistant most firms already owned. Two of the top four products carry no legal branding at all.
Around the same time, Yath I., Chief Transformation Officer at Morgan & Morgan, P.A., published a piece arguing the whole question is backwards. The first question is not which AI tool to buy. The first question is how your organization should change once the technology exists. Read the survey and his argument together and a clear plan for a personal injury firm comes into focus. Here is my read.
What the ILTA 2026 Technology Survey Found
The survey covered more than 500 law firms across a dozen countries, representing about 140,000 lawyers. Ninety-four percent of those firms report using or exploring generative AI, up from 80 percent a year ago. Among firms with fewer than 50 lawyers, the figure is 85 percent.
When ILTA asked which tools firms are using or exploring, Microsoft 365 Copilot led at 76 percent. Second place was a three-way tie: Thomson Reuters CoCounsel at 44 percent, Anthropic’s Claude at 44 percent and Harvey at 43 percent. Among firms with fewer than 50 lawyers, Copilot still leads at 59 percent, followed by Claude at 42 percent and CoCounsel at 40 percent. Harvey drops to 22 percent in the small-firm segment and Legora to 6 percent.
Saying your firm uses AI and putting a tool in every lawyer’s hands are different things. Only two products reached full deployment among lawyers at half or more of responding firms: Microsoft Copilot at 52 percent and Westlaw Advantage at 50 percent. Bob Ambrogi of LawSites flagged a problem here. The Copilot leading the deployment chart appears to be the free version bundled with Windows, and the 76 percent usage figure belongs to the paid Microsoft 365 product. Read the deployment number with some skepticism.
The following chart from Bob Ambrogi tells the story:
The piloting numbers show where firms are testing next. The two most-tested products both come from iManage, the document management company: iManage AI Enrichment at 91 percent and Ask iManage at 80 percent. Claude follows at 68 percent, then Litera One at 64 percent, NetDocuments’ ndMAX App Builder at 58 percent and Clio’s Vincent AI at 56 percent. Firms are testing AI inside the systems where their documents and matters already live.
How are firms using AI? The top four uses all landed between 73 and 79 percent: first drafts of documents, meeting notes and summaries, legal research, and summarizing or analyzing long documents. Generative AI also appeared on ILTA’s list of top security challenges for the first time and debuted in second place, behind only user behavior. Seventy-eight percent of responding firms report having a formal policy for generative AI use.
The Survey Answers the Wrong Question First
Every number above answers the same question: which tool. Ithayakumar’s point is the tool decision comes second. Morgan & Morgan grew from $250 million to $2.5 billion in revenue over ten years, by his account, and the firm did not get there by buying software. The firm reorganized around technology before the technology was fully in place, and the firm is doing the same with AI now. His line sums up the approach: your org chart is part of your AI strategy.
The survey supports him in a way he did not intend. Copilot wins because Copilot sits inside Word, Outlook and Teams, where staff already spend the day. Nobody changed a habit to try the product. The tools winning pilots are the ones inside document management systems. Firms are not choosing the smartest model. They are choosing the one closest to existing work. If the work itself is poorly organized, the AI inherits the disorganization. Fix the work first.
Audit Tasks, Not Roles
Ithayakumar’s first step is a simple one but frequently overlooked. Before you evaluate a single tool, list every responsibility your paralegal handles in a week. Do the same for case managers, legal assistants and attorneys. Then sort the list into three buckets: tasks AI handles today with a quality check, tasks requiring your firm’s legal judgment, and tasks requiring specialized knowledge your firm does not need to own.
The first bucket is where general AI belongs: records summaries, meeting notes, intake forms, mail sorting, first drafts of routine correspondence. The survey shows firms getting results here, and most of your team already has access through Microsoft 365 or your case management vendor.
The second bucket is the work only your lawyers and paralegals should do: client counseling, medical chronologies, discovery, demand strategy, trial preparation. Ithayakumar lists several places to redeploy recovered hours, and the list reads like a PI case value checklist: more thorough client communication, chasing key medical documentation, searching harder for additional coverage, and identifying liens early.
Lien work is where I part ways with him on one detail. Identifying liens early matters. Resolving them is a different task, and the resolution work belongs in the third bucket.
The Work in the Third Bucket
A single case file often carries several potential lienholders: Medicare, Medicaid, a Medicare Advantage plan, an ERISA plan, the VA, TRICARE, hospitals and individual providers. Someone has to identify each one, verify coverage, request the claim itemization, audit the charges for unrelated treatment, negotiate a reduction and secure a final demand in writing before disbursement. At most firms, the someone is a paralegal who also handles client calls, records requests and discovery deadlines. Hours on hold with a recovery contractor add nothing to case value.
None of the products in the ILTA survey touch this work. A general model will summarize 600 pages of medical records in minutes. The same model has no way to confirm whether your client’s health plan is a self-funded ERISA plan, whether a Medicare Advantage plan paid accident-related claims, or whether a hospital perfected its lien under your state’s statute. Answering those questions takes payer data, payer contacts and people who do the work every day. The job is part technology and part specialized labor, which is why I built Synergy the way I did.
Synergy runs the work on VerifiAI, our own AI platform for lien identification and verification, and pairs the technology with a team of lien resolution specialists who handle the full process from first inquiry through final resolution. Your staff sends the case. VerifiAI flags the likely lienholders, including the Medicare Advantage and ERISA plans firms miss most often. Our specialists confirm what is owed, dispute unrelated charges, negotiate the reduction and deliver a final demand in writing before you disburse. Your paralegals get those hours back for the second bucket.
Ithayakumar makes the same point from a different direction when he describes separating Morgan & Morgan’s pre-suit department from litigation. Case-level data showed the firm’s best litigators spending most of their time on pre-suit work. The fix was structural, not technical. Separate the functions, and the best people return to the work with the highest return. Moving lien resolution out of the paralegal’s job description is the same fix at a smaller scale.
Change the Scorecard With the Work
Here is the part most firms skip. If AI now produces the first draft of every demand, paying case managers a bonus on demand volume rewards the wrong thing. Ithayakumar’s answer is to redefine performance around the work the technology freed people to do: client communication and satisfaction, the number of matters a person manages well, and the activities with a direct effect on case value.
Apply the same logic to lien work. A paralegal who no longer spends eight hours a week on hold with a recovery contractor should be measured on something else. Count client touches per file. Track days from treatment completion to demand. Track coverage found after intake. Reward those numbers, and the recovered time turns into fee revenue instead of slack.
A 10-person firm with three people in pre-litigation, three in litigation and two on closings does not get smaller under this plan. The same team handles more matters or more complex matters without burning out. This is the growth path for a firm competing against larger firms with bigger budgets.
End the Pilot Loop
ILTA’s piloting numbers show a lot of testing. Ithayakumar’s warning is the pilot loop: firms running trial after trial while waiting for the perfect tool. You do not need to be an expert on every product. You need to be an expert on your own operations. The worst technology decision is no decision.
Two guardrails before you commit. First, write the AI policy. If 78 percent of ILTA firms have a written generative AI policy and your firm does not, start there. Name the approved tools. State which client data goes into which tool. Require business or enterprise versions with contract terms barring the vendor from training on your data. Require human review of every AI output before filing or sending. Apply the same standard to any vendor touching protected health information, and confirm a business associate agreement is in place.
Second, run one pilot with a clock on the task. Medical record summaries are a good candidate. Time the work before and after, and have a lawyer check accuracy on every file. Then decide and move on.
What to Do This Quarter
- Build the task inventory. Every role, every weekly responsibility, sorted into the three buckets. Price the hours in each bucket.
- Inventory the AI you already pay for. Your Microsoft 365 subscription likely includes some version of Copilot, and your research and document management vendors have likely added AI features since your last renewal.
- Write or update the AI policy and train every staff member on the policy before any pilot starts.
- Move the third bucket out of the building. Lien identification, verification and resolution will sit near the top of the list. Compare the hourly cost against handing the work to a specialist and redeploying your team.
- Rewrite the scorecard. For each role, replace at least one output metric the technology now handles with a metric tied to client service or case value.
The survey tells you which tools firms are buying. Ithayakumar tells you the purchase matters less than the organization around the purchase. Both are right. Decide which work belongs with your lawyers and legal staff, which belongs with software and which belongs with a specialist, and build the firm around the answer.
Which task would you move out of your paralegals’ week first? If lien resolution is on your list, my team at Synergy will walk you through what the handoff looks like on your own files.
Why Synergy Is Built for This Moment
At Synergy, we’ve always believed that lawyers should focus on securing justice while we handle the friction points that slow firms down. Now, with technology creating competitive pressures and the need to be as efficient as possible, we’re doubling down on that mission, helping firms integrate the best tools with the best people to achieve Peak Practice.
🔗 Want more insights like this?
If you’re a personal injury lawyer ready to scale, streamline, and step into your role as CEO, let’s talk. Join the Peak Practice Community, and learn how Synergy can help you eliminate settlement bottlenecks, resolve complex liens, and maximize recoveries. Learn more here: https://partnerwithsynergy.com/peak-practice/
If you want to grow and scale your law firm more effectively, consider partnering with Synergy for lien resolution. Learn more at: https://partnerwithsynergy.com/liens/