The Accountability Challenge in Claims Adjustment

When an AI system denies a claim, who answers for it? That question sits at the heart of the ethical debate over AI in claims adjustment — and it's the one reason full replacement isn't happening anytime soon.

The core problem is simple: current AI cannot take legal or professional responsibility for investigating a loss, interpreting coverage, or making and defending the final claim decision. As Jim Cresse, a National General Adjuster with more than 35 years of experience, puts it: "AI can suggest. The adjuster must decide." Claims handling is a highly regulated activity, and the accountability gap between a machine's output and a defensible, documented decision remains unbridgeable.

The industry's own data confirms where the line sits. Automation is currently used in claim assignment, but not in the decision-making process for claim settlements or denials, according to a 2026 state of claims automation report. The judgment calls that determine what a policyholder receives still rest with humans.

Regulators are formalizing that requirement. The EU AI Act classifies AI used in insurance claims as high-risk, mandating documented decision logic, human oversight mechanisms, and audit trails for every automated decision, according to compliance analysis. In the US, 25 jurisdictions have adopted the NAIC AI model bulletin as of April 2026, and Texas regulators have clarified that AI-enabled insurance activities remain subject to existing insurance law.

Yet governance is lagging adoption. While 70% of organizations have AI risk committees, only 14% report being operationally prepared to implement AI safely, per Sedgwick's 2026 Loss Adjusting Insights Report. That gap raises real ethical concerns about how AI is being deployed today.

Unresolved accountability questions include:

• Under what circumstances, if ever, can a machine approve or deny a claim?
• Who bears responsibility when an AI-assisted decision is wrong — the insurer, the vendor, or the supervising adjuster?
• How can policyholders appeal a decision made by a system they cannot interrogate?

There's also an implementation-ethics problem inside carriers. 98% of Glassdoor reviews from claims adjusters mentioning AI were negative — the highest skepticism of any insurance role — with adjusters describing 15–20 minutes spent correcting AI-generated mistakes that supervisors never verify, according to Glassdoor/Indeed research reported by Insurance Business. AI deployed without human verification doesn't just create rework; it erodes trust in the decisions themselves.

The takeaway: accountability, not capability, is the barrier. Until regulators, courts, and insurers resolve who answers for a machine's decision, a licensed human will remain the one signing off — and the technology's role will be to support that person, not replace them. At Agents by AIQ, we see the same principle apply across insurance operations: AI handles the busywork, humans own the decisions.

The Impact of AI on the Claims Adjustment Workforce

The numbers paint a paradox: insurers are investing heavily in automation, yet the people who understand claims are quietly disappearing. Entry-level adjuster postings have fallen roughly 50% since early 2024 — far steeper than the 15% decline across entry-level jobs overall — while senior-level postings remain about 80% above 2017 levels, according to industry analysis.

That gap matters because of how adjusters have always been made. Junior staff learn by handling routine claims, then graduate to complex losses. If AI absorbs the routine work that once served as on-the-job training, experienced adjusters stop being built. As the trade press framing puts it: if junior roles disappear, experienced adjusters stop being built.

The timing could hardly be worse. Nearly 25% of claims adjusters are expected to retire by the end of 2027, and in Sedgwick's 2026 Loss Adjusting Insights Report, 73% of insurance professionals said the greatest impact of that retirement wave will be lost institutional knowledge, while 62% expect a mentorship shortage. Strip out the entry-level rung and the pipeline problem compounds: there will be no one to inherit what retiring experts know.

This is where the ethical questions get concrete:

• Who bears responsibility when AI-assisted training gaps produce under-qualified adjusters handling complex losses?
• How do firms document and preserve institutional knowledge before it walks out the door?
• What obligation do insurers have to workers whose roles are being reshaped without their input?

Worker sentiment suggests that obligation is already being mishandled. Glassdoor and Indeed research found that 98% of claims adjuster reviews mentioning AI were negative — the highest skepticism of any insurance role. One adjuster described spending 15–20 minutes correcting AI-generated mistakes because supervisors don't verify anything. That's an implementation-ethics problem: AI deployed without human verification creates rework and erodes trust.

Important caveats apply. The data does not prove AI is causing these hiring declines — cost pressures, shifting claim volumes, and a broader slowdown may contribute. And 49% of carriers still expect to increase staff over the next year, per an Aon and Jacobson Group labor study.

The practical takeaway for firms: automation can handle busywork, but someone still has to own the judgment calls — and that someone needs a pipeline behind them. At Agents by AIQ, we see the same principle with agencies and small firms adopting AI for phones and follow-up: the technology works best when it supports people, not when it quietly replaces the training ground they need.

Implementing AI in Claims Adjustment Responsibly

AI is reshaping claims adjustment, but its implementation demands careful balance to preserve human expertise and ethical integrity. As insurers adopt AI for tasks like document analysis and fraud detection, the need for human oversight remains critical to ensure accountability and maintain the expertise pipeline.

A 2026 report highlights that AI automation has reduced claims resolution time by 75% in some cases, yet final decisions—such as settlement approvals or denials—still require human judgment. Hesper AI’s data underscores that automation is limited to claim assignment, not adjudication, emphasizing the irreplaceable role of adjusters in high-stakes scenarios.

To implement AI responsibly, organizations must prioritize:

• Establish human oversight mechanisms, ensuring adjusters review AI-generated recommendations, especially for complex or disputed claims.
• Adhere to regulatory frameworks like the EU AI Act, which mandates transparent decision logic and audit trails for high-risk applications.
• Invest in training to help adjusters adapt to AI tools, addressing concerns about skill gaps and the erosion of entry-level training opportunities.

Ethical challenges also arise from AI’s impact on workforce dynamics. Research reveals 98% of adjuster reviews mentioning AI were negative, citing frustration over errors and lack of supervisory verification. This underscores the need for systems that empower, rather than undermine, human expertise.

AI agents can alleviate administrative burdens, handling tasks like phone answering, lead follow-up, and data entry. For businesses seeking to streamline operations, Agents by AIQ offers tailored solutions to manage routine work, allowing teams to focus on strategic, human-centric responsibilities.

By integrating AI as a supportive tool rather than a replacement, insurers can enhance efficiency while upholding ethical standards and preserving the skills needed for complex claims.