By Reid Zeising, Co-founder of GAIN Servicing
Some firms and practices in this industry have spent the past year working with AI tools on their real caseloads. They have tested a chronology generator against a messy medical file, watched a case management system flag the wrong thing, corrected it, and tried again. Others are still evaluating, waiting for the tools to feel more proven before they commit real cases to them. That instinct feels responsible. It’s also the reason the gap between those two groups is widening every month, and it has very little to do with which software either one eventually chooses.
The mistake is assuming the advantage sits in the technology itself, as if the firm that picks the best platform wins. What actually separates these two groups is something that cannot be purchased. It is the accumulated experience of feeding a system real cases, seeing where it gets the chronology wrong or misses a lien, and learning exactly what it needs from you to do better next time. That skill only comes from using these tools on live matters, under real-time pressure, with real documents. A firm cannot read its way into that knowledge, and it cannot buy it fully built from a vendor. It has to build it through repetition, and every month spent waiting is a month that knowledge does not get built.
Where This Plays Out in Real Cases
For attorneys, this plays out in how a practice uses the research and drafting tools now available for personal injury work. Roughly 79% of legal professionals now report using AI in some form, and firms with wide adoption are considerably more likely to report real revenue growth. That gap is not about access. Every firm can license similar tools today. It’s about the attorneys and staff who have spent months learning which chronology tool actually catches a treatment gap and which one just organizes what was already there, versus the firms still deciding whether to start. The skill of knowing what to verify, what to trust and what a draft is missing does not transfer through a vendor demo.
For providers, the same pattern shows up in how a practice manages documentation and RCM. AI-supported systems can flag an aging receivable or a record that looks incomplete, but a billing team only learns to trust and correctly interpret those flags by working through real claims and real denials over time. What separates providers who use these tools well has little to do with which system they run. It comes down to how many cycles they have logged with it, enough to know when a flag actually matters and when it does not. That kind of judgment is exactly what industry-wide reporting on case management adoption shows is still uneven across the field, even as more firms move onto these platforms every year.
What Waiting Actually Costs
The instinct to wait comes from a reasonable place. Personal injury work touches medical records, liens and client funds, and a flawed output in that environment does not stay contained. It moves into a demand letter, a settlement calculation or a client’s expectations. That risk is real, and it is exactly why early use of these tools should happen with attorney and staff review built into every step.
But waiting does not remove that risk. It shifts it into a different form. While some practices spend another quarter deciding, others are already past the early mistakes, already training their staff on what these systems get wrong, and already moving cases faster because of it. The practices that engaged early are not learning the basics anymore. They are refining. The ones still on the sidelines will eventually adopt the same tools, but they will do it from further behind, competing for the same clients and the same talent against firms that already know how to use the technology well.
What I have found actually works is treating this as a discipline you build inside live work. Start with the single bottleneck creating the most friction, whether that is record review, intake or lien tracking, put review steps around it, and let your team learn from real cases instead of a sandbox. That approach builds the judgment that actually compounds. The technology available to providers and attorneys today will keep changing. The head start belongs to whoever has already spent the time learning how to direct it well, and that head start does not wait for anyone else to catch up.
For a broader business perspective on this, I break it down in my recent article in Entrepreneur: https://www.entrepreneur.com/leadership/the-leaders-who-engaged-with-ai-early-are-no-longer/504588.