Insurance AI Isn’t Going Anywhere: What PI Lawyers Must Know in 2025

Insurance AI Isn’t Going Anywhere: What PI Lawyers Must Know in 2025

Since the early 1990s, insurers have invested heavily in AI-driven claims evaluation software designed to do one thing: minimize payouts. Tools like Colossus, Liability Navigator (L-Nav), and ClaimIQ have become deeply embedded in claims handling across the U.S., allowing carriers to streamline operations and control settlement costs.

But here’s the hard truth for plaintiff-side attorneys:

Insurance AI isn’t going away. It isn’t being banned. And despite increasing public scrutiny, it isn’t being meaningfully regulated.

Understanding how these systems work is no longer optional. If you’re not factoring insurance AI into your settlement strategy, you’re negotiating at a disadvantage. In this article, we’ll explore:

  • How insurers use AI to undervalue claims;

  • Why legislative efforts to regulate these systems have largely failed;

  • What this means for personal injury attorneys negotiating with carriers in 2025;

  • How Settlement Intelligence’s AI-optimized demand letters help you fight back and win.

The Origins of Insurance AI: Colossus and the Systematic Lowball

AI in insurance claims handling didn’t start with ChatGPT, it started with Colossus. Introduced in the early 1990s, Colossus promised carriers a standardized, data-driven approach to bodily injury valuation. Behind the curtain, however, bad faith litigators quickly uncovered its real purpose:

  • Undervaluing personal injury claims by categorizing injuries into rigid “value bands”

  • Using historic claims data, weighted in insurers' favor to control payouts.

  • Encouraging adjusters to rely on algorithmic “recommendations” rather than independent analysis.

In 2010, after consumer complaints and multi-state investigations, the California Department of Insurance fined Allstate $10 million and required the company to disclose that Colossus was merely a “tool for evaluation,” not the sole determinant of claims value (Insurance Journal, 2010).

But despite the headlines, the sanction was a slap on the wrist. Colossus remained in widespread use. And over the years, it evolved into a network of modern, interconnected AI-driven tools now used by nearly every major carrier.

For a deeper dive into Colossus’ methodology and its impact on settlement offers, read our post: Where Plaintiff Advocacy Meets Insurer Logic: The AI-Ready Demand Letter.


The Modern Insurer Toolkit: Colossus, Liability Navigator, ClaimIQ & Beyond

Today, Colossus is just one part of a sophisticated AI ecosystem that evaluates claims, predicts verdicts, and informs settlement ranges:

  • Colossus — Evaluates bodily injury severity and treatment types based on more than 12,000 factors and jurisdictional data to assign a value range.

  • Liability Navigator — Used by adjusters to calculate probability-weighted liability based on accident facts.

  • ClaimIQ — Compares claims against a massive internal database of verdicts and settlements to benchmark value bands

These systems are powerful, and opaque. Carriers rarely disclose their use, but personal injury attorneys see the impact every day: lowball offers that don’t reflect the full value of a claim.

Regulatory Attempts: Why Insurer AI Remains Unchecked

Public awareness of AI’s role in claims handling is growing, and policymakers are starting to act. But the legal landscape remains fragmented, and efforts to regulate insurer AI have largely fallen short.

Florida’s SB 794: Well-Intentioned, Dead on Arrival

In early 2025, Florida lawmakers introduced SB 794, a bill aimed squarely at accountability in AI-driven claim denials.

The bill would have:

  • Required licensed human review for any claim denial (full or partial).

  • Mandated that denial letters identify the reviewer and confirm that AI wasn’t the sole decision-maker.

  • Empowered Florida’s Office of Insurance Regulation to audit insurer compliance.

Outcome: After unanimous support in the Banking & Insurance Committee, SB 794 died in appropriations without reaching the governor’s desk (Florida Senate Bill 794).

For Florida consumers and plaintiff lawyers, nothing has changed. Insurers can continue using AI like Colossus to shape valuations, without disclosing its role.

California’s SB 1120: A Narrow Win

California has been more aggressive. Its Physicians Make Decisions Act (SB 1120), effective January 1, 2025, requires that licensed professionals — not AI — review medical necessity denials.

But SB 1120 doesn’t cover settlement valuation or bodily injury claims. Colossus and ClaimIQ remain untouched.

NAIC’s AI Model Bulletin: Non-Binding Guidance

In December 2023, the National Association of Insurance Commissioners (NAIC) issued its Model Bulletin on AI, urging carriers to ensure:

  • Transparency in AI usage,

  • Protections against algorithmic bias,

  • Human oversight of automated decisions.

So far, 23 states have adopted the bulletin, but because its guidance and not law, insurers face no penalties for ignoring it.

Litigation Trends: Cigna, UnitedHealthcare & More

Regulators aren’t the only ones paying attention. Plaintiffs have started fighting back:

  • Cigna faces a class-action lawsuit alleging its algorithm denied thousands of claims in bulk without human review.

  • UnitedHealthcare is being sued over Medicare Advantage denials allegedly based on AI systems rather than licensed physician determinations.

These cases highlight a growing risk for carriers, but they’re early-stage battles, and outcomes are far from certain.

What This Means for Personal Injury Lawyers

With no meaningful regulation limiting insurer AI in claims valuation, plaintiff lawyers must adapt. Here’s what you can do now:

1. Understand the Systems Behind the Numbers

Insurers rely on AI to set settlement bands, analyze attorney performance, and predict verdict ranges. If you don’t know how tools like Colossus work, you’re negotiating blind.

Action Step: Build AI literacy within your firm. Train your team to recognize lowball offers influenced by algorithmic “value caps,” and prepare every claim to include evidence that is needed to trigger insurance AI. 

2. Target AI Usage in Discovery

AI introduces vulnerabilities for insurers:

  • Rigid reliance on outdated data;

  • Failure to consider unique client factors;

  • Instances where AI outputs override human adjuster discretion.

Craft discovery requests that demand:

  • Disclosure of what AI tools were used,

  • Inputs fed into those systems,

  • Documentation of any human override (or lack thereof).

These tactics create leverage and potential grounds for bad faith claims.

3. Trigger Insurance AI: Better Demand Letters to Fight Back

This is where Settlement Intelligence gives plaintiff lawyers a critical edge.

Our patent-pending platform produces AI-optimized demand letters designed to interface directly with insurer evaluation software like Colossus, Liability Navigator, and ClaimIQ.

By structuring demands to “speak the same language” as these systems, our technology ensures:

  • Injuries are categorized for maximum valuation;

  • Special damages are highlighted in algorithm-friendly formats;

  • Liability facts are framed in a way that drives higher offers.

For more on this strategy, see: Where Plaintiff Advocacy Meets Insurer Logic: The AI Ready Demand Letter

The Bottom Line

Insurer AI is here to stay, and plaintiff lawyers must adapt. Despite isolated legislative wins, there’s still:

  • No federal regulation governing insurance AI.

  • Few state-level protections, and none addressing settlement valuations.

  • Rapid expansion of AI tools designed to minimize payouts.

Understanding how Colossus and similar systems shape your client’s settlement offer isn’t optional anymore, it’s a strategic imperative.

At Settlement Intelligence, we help PI lawyers maximize settlement offers by delivering AI-driven demand letters that give you leverage against insurers’ technology.

Subscribe today at app.demandletters.ai/signup

Back to blog