Tainton: The Right AI Doesn't Just Make Claims Professionals Faster, It Makes Them Better

Wednesday, March 25, 2026 | 0

Earlier this year, I was on a panel at the California DWC annual Educational Conference in Oakland and then again in Los Angeles, a session on the state of AI in workers’ compensation. Those sessions brought together clinical, legal, technical and operational perspectives, and that range made it one of the richer conversations I’ve been part of.

The sessions moved through questions that don’t get raised often enough: What happens when an AI-generated summary becomes the only record decision-makers actually review, and what critical information gets lost in that compression? Is AI-driven triage locking claims into paths before enough information is in? When AI errors compound at scale, who is accountable and how reversible is the damage? What judgment should never be delegated, regardless of how good the model is?

The theme I kept coming back to was this: The right technology doesn’t just make claims professionals faster. It makes them better.

That same thread ran through a presentation I gave at PropertyCasualty360’s Complex Claims and Litigation Forum in March, which approached the problem from the other end of the lifecycle. On the litigation side, much of what drives bodily injury disputes comes down to how medical evidence is interpreted. The question isn’t whether AI can accelerate that process. It can. The question is whether the output holds up when it matters: in depositions, in discovery and under cross-examination. Speed without defensibility isn’t a solution.

Both conversations pointed to the same gap: Claims professionals across the team (adjusters, nurses, attorneys, paralegals, and others) spend roughly half their time organizing and reading documents before making a single decision. AI that simply moves faster doesn’t close that gap; it accelerates it. When AI surfaces the right information at the right moment with reasoning that is visible and traceable, claims professionals make decisions with more confidence, more context and fewer blind spots.

That infrastructure, as I see it, rests on four pillars.

The Medical Record Problem

Workers’ compensation is one of the most document-intensive lines in insurance. A single claim can generate hundreds of pages of medical records: physician notes, diagnostic reports, pharmacy records, and IME opinions, plus surveillance documentation, legal filings, and more often, duplicated, misfiled and arriving out of sequence. The question I ask organizations isn’t whether they have a document problem. It’s what that problem is actually costing them: late reserve adjustments, missed treatment guideline deviations and delayed return-to-work coordination. Most of it traces back to the same upstream failure: unstructured data entering the claims process without being made usable. Data quality upstream determines everything downstream.

Where AI is Actually Delivering, and Where New Risks Emerge

AI has delivered real value in workers’ comp: automated medical chronologies, risk signals surfaced earlier in the claims lifecycle and treatment pattern analysis run against accepted clinical standards before the first utilization review decision is made. But AI-driven triage carries a risk the DWC panel surfaced clearly: locking claims into paths too early. When an AI-generated summary becomes the only document decision-makers review, you lose the nuance that often matters most: the atypical injury, the comorbidity or the claim that doesn’t fit the model’s training. AI that directs attention to the right signal inside a file sharpens professional judgment. AI that replaces the file quietly degrades it.

Defensibility and Documentation

This is where the Complex Claims and Litigation conversation connects directly to workers’ comp. Nuclear verdicts and social inflation are reshaping exposure, and as AI-assisted claim decisions become more common, what happens when those decisions reach litigation is no longer hypothetical. Defense counsel needs to reconstruct the basis for a claim decision: what was reviewed, what flags were generated and how human judgment interacted with AI output at each stage. Explainability isn’t a nice-to-have; it's the product. An AI system that produces a risk finding without a traceable path back to the underlying records isn’t just technically incomplete; it's the kind of inconsistency that plaintiff counsel depends on. Auditable documentation that withstands discovery scrutiny is the standard.

The Future Claims Professional

Five years from now, the claims professionals gaining the most ground will be the ones who learned to use AI as a thinking partner, not a shortcut. The judgment, the relationships and the ability to read a complicated situation — none of that goes away. Better information sharpens all of it. What changes is the baseline: claims professionals who no longer spend the first part of their day sorting a raw file, who get risk signals before a claim escalates rather than after and who can ask a question about a complex file and get a cited, traceable answer in seconds. Return-to-work outcomes remain the clearest signal that a claims process is actually working. When the intelligence infrastructure is doing its job, that signal shows up earlier and more consistently.

The best conversations at both the DWC annual Educational Conference and Complex Claims and Litigation Forum weren’t about technology. They were about outcomes: faster time to first reserve decision, earlier identification of claims at risk for prolonged duration and better-coordinated defense before costs mount. What claims organizations can control is the quality of the intelligence they’re working with.

Workers’ compensation has always rewarded early action. The profession has the standards, the data, and the expertise. What it needs is the infrastructure to act on that knowledge, not just faster but better, before the cost of delay compounds.

Mark Tainton is SVP of Data Solutions at Wisedocs, the decision intelligence platform for insurance claims. He recently presented at PropertyCasutly360’s Complex Claims and Litigation Forum and spoke at the California DWC Annual Educational Conference on AI, accountability, and the future of decision-making in workers’ compensation.