A law firm turned weeks of research into a 7-minute summary. Most solo attorneys aren't close.

The result

Laffey Bucci D'Andrea Reich & Ryan is a small firm that represents abuse survivors (document-heavy, deadline-driven work built on medical records, depositions, and long factual timelines. Using Westlaw's AI-assisted research and CoCounsel's summarization tools, attorney Guy D'Andrea cut legal research time by up to 80%. Work that used to take his team weeks) building a clean summary and timeline from a stack of records, now takes under seven minutes.

"I'll have a summary that's near perfect and a timeline . . . in less than seven minutes. It saves us weeks.". Guy D'Andrea

A few states over, solo attorney Safa Riadh of Valiant Law used the same category of tool to do something more interesting than save time: he took on case types he used to turn away because the research overhead wasn't worth it. Faster research didn't just compress his week, it changed what kind of law he could afford to practice.

Not a fluke, the pattern shows up at scale

Thomson Reuters profiled five small and midsize firms using CoCounsel and found the same shape repeating: 63% faster document review and drafting, attorneys finding twice as many relevant cases in the same amount of time, and up to 12 hours a week freed per attorney, enough for roughly 10% more matter capacity with no added headcount.

A separate Forrester study (commissioned by Thomson Reuters) modeled a composite 500-attorney firm adopting the same category of tool and found a 400% three-year ROI, driven by $20.3M in incremental revenue from added matter capacity rather than headcount cuts. That's a large-firm model, not small-firm data, but it says the economics of "verified AI research done properly" hold up at every size, not just the anecdote size.

The other half of the same coin

The same underlying technology, pointed at the same task, produces the opposite outcome when nothing checks its work. A federal judge in Oregon sanctioned two attorneys $110,000 in May 2026 for 23 fabricated AI-generated citations (the largest AI-hallucination penalty on record. A public tracker lists roughly 1,490 court decisions worldwide where a party relied on AI-hallucinated material, and 90% of the attorneys sanctioned for it were solo or small-firm practitioners) the same buyer as the case study above.

The difference between the two outcomes was never "did they use AI." Both groups did. The difference was whether the research ran through a workflow with a verification step, or through a chat window with none.

What actually separates the two groups

It isn't firm size, budget, or which model they subscribed to. It's whether AI is wired into a workflow (trained on the firm's own matters, with citations checked against a real source before they reach a filing) or used the way most solo attorneys use it today: a faster search bar, one-off, no guardrail, no habit built around it. That gap is exactly what an audit is for: finding which two or three tasks in a practice are worth automating first, and building the verification in from day one rather than after a sanction.

Sources: Thomson Reuters Legal, "ROI Reality check: 5 small and mid sized law firms using AI" (Feb 2026); Thomson Reuters / Forrester Consulting, Total Economic Impact™ study of CoCounsel Legal (2026); Fortune, "Would you hire the lawyer who just got sanctioned for using AI?" (May 2026); Damien Charlotin, AI Hallucination Cases Database.