← Back to News
ANALYSIS

Touchless Auto Claims Cross 60% at Top US Carriers — The Quiet AI Transition Reshaping Insurance

GEICO, Progressive, State Farm, Allstate, and USAA are now settling more than 60% of US auto physical damage claims without any human adjuster touching the file. Tractable, CCC, Hi Marley, and Snapsheet have built the vision-model infrastructure that quietly displaced one of the largest field-based white-collar workforces in the country.

By Michael Eakins•• min read
InsuranceClaims AIComputer VisionTractableCCC Intelligent SolutionsWorkforce DisplacementTouchless ClaimsAuto InsuranceProperty CasualtyAI Vision Models

The Insurance Industry Just Crossed a Quiet Threshold

Across the top five US personal auto insurance carriers — GEICO, Progressive, State Farm, Allstate, and USAA — more than 60% of auto physical damage claims now settle without any human claims employee touching the file. The First Notice of Loss arrives, the policyholder uploads photographs through the carrier's mobile app, an AI vision model identifies and assesses the damage, a parts and labor estimate is generated against real-time pricing data, a settlement offer is produced, and on acceptance the payment is initiated within the hour.

This is the touchless claims threshold the industry has been moving toward since 2019. It crossed 50% at the leading carriers in early 2025. It crossed 60% at the median in early 2026. By 2028 it will cross 80%. The shape of the curve is now locked in.

The trajectory has substantial implications for the approximately 320,000 US claims adjusters, examiners, and investigators whose work the AI is replacing — but the industry conversation has been dominated by the customer experience and unit economics story rather than the workforce story. Both deserve attention.

What Changed and When

The technical breakthrough that enabled the touchless threshold was not a single innovation. It was the convergence of several maturations:

Vision model accuracy on common auto damage types crossed the human adjuster baseline in 2023, with frontier vision models trained on tens of millions of labeled damage examples now matching or exceeding the average certified adjuster on the same task.

Mobile photo capture matured, with structured workflows that guide the policyholder through required angles, real-time quality control, and automated re-prompts for unusable images. The 2026 generation of mobile claim apps captures photos of sufficient quality for AI assessment in roughly 90% of first attempts.

Real-time parts pricing integration through Mitchell, CCC, and Audatex APIs allowed AI estimates to reflect current parts prices accurately, which had been a credibility barrier for AI estimates as late as 2022.

Instant payment rails through Mastercard Send, Visa Direct, and ACH Now enabled same-hour settlement disbursement, which materially improved the customer experience case for AI-driven claims.

Carrier comfort with AI decision liability matured through 2024- 2025 as state insurance commissioners issued AI guidance frameworks that clarified compliance expectations without prohibiting AI use.

The combined effect is that the touchless claims threshold passed through 50% in early 2025 and 60% in early 2026 with surprisingly little public commentary outside trade press.

The Vendor Picture

The competitive structure of AI claims processing in 2026 has stabilized into five-vendor competition with distinct positioning.

Tractable has been the most aggressive growth story, building from a vision-model-first approach to auto damage and expanding into property and salvage. Recently disclosed that more than 60% of US auto claims at top-five carriers flow through Tractable's vision pipeline at some stage.

CCC Intelligent Solutions is the legacy market leader, with deep network effects from the CCC ONE platform that virtually every collision repair shop already uses. Has been less aggressive on displacement than the pure-play vendors but has the deepest installed base.

Hi Marley has built strength in customer communications and the policyholder experience side of claims processing, with strong adoption among mid-market carriers where customer experience differentiation matters most.

Snapsheet pioneered virtual appraisal — the policyholder-photographs-her-own-damage workflow that is now industry standard. Reports the highest touchless rate of any major vendor in publicly disclosed figures.

Solera Audatex competes internationally with CCC and is growing in US auto claims, with significant investment in AI capabilities to match Tractable's growth.

Underneath all of these vendors sits the foundation model layer. Vision pipelines increasingly build on multimodal frontier models from OpenAI, Anthropic, and Google. The vendor differentiation is narrowing toward training data assets, carrier integration breadth, and workflow sophistication rather than raw vision capability.

The Economics Are Overwhelming

The financial case for AI claims processing is among the strongest in all of enterprise AI. Fully-loaded human adjuster cost is approximately $51 per claim. AI claims processing vendor charges run $8-$18 per claim at scale. Even before accounting for the value of faster settlements (improving customer retention by 4-7%), reduced claim leakage (3-7% of paid claim value), and reduced reinspection and supplemental cycles, the labor cost arbitrage alone justifies deployment at virtually any carrier above $1B in annual claims volume.

Once the secondary benefits are factored in, AI claims processing has negative net cost per claim at the carrier — generating more incremental margin than it costs. This is rare in enterprise AI unit economics and explains why deployment velocity at major carriers is bounded by vendor implementation capacity rather than by buyer willingness.

The implications for combined ratios are substantial. A 3% reduction in claim leakage translates to roughly 1.5-2 percentage point improvement in combined ratio at a typical personal lines carrier, which is an enormous shift in underwriting profitability.

The Workforce Picture Nobody Is Discussing

The implied math on the workforce is unavoidable. If touchless rates continue on the current trajectory — roughly +10 percentage points per year through 2028, then plateauing in the high 80s through the early 2030s — the US claims adjuster workforce will contract from approximately 320,000 in 2024 to roughly 95,000 by 2032.

CrashBytes's detailed analysis of the AI claims processing displacement curve and 290,000 adjuster jobs disappearing by 2032 projects the same trajectory across most personal lines and progressive extension into property, workers compensation, and bodily injury claims.

The contraction is back-loaded for two reasons. First, the largest carriers absorbed initial deployment through attrition rather than forcing layoffs. Second, the major carriers have run AI claims processing in parallel with human adjusters for 6-12 months at each deployment phase before crossing the threshold to material headcount reduction.

But the trajectory is now mechanical. Carrier CFOs have committed to the deployment, the unit economics have validated, the regulatory environment has clarified, and the residual technical risk is manageable. The workforce contraction will play out over the next seven years on the timeline implied by the touchless rate curve.

The Body Shop Network Effect

One under-discussed implication is the secondary effect on the collision repair industry. The elimination of in-person inspection by adjusters reduces the relationship value of the local adjuster to the body shop. Body shops that integrate well with carrier AI workflows — providing structured photographic documentation, accepting AI-generated estimates without supplemental disputes, completing repairs within AI-projected cycle times — receive disproportionately more referrals.

The collision repair industry is consolidating partly in response. The major networks (Caliber Collision, Service King, Crash Champions, Gerber Collision) are positioning as "AI-friendly" repair partners with the carriers, with standardized workflows and tight integration. Independent shops that cannot match this capability are losing market share.

The aggregate secondary employment effect on roughly 200,000 body shop technicians and managers is real but harder to quantify. Network consolidation does not necessarily reduce headcount as much as it reshuffles which employer captures the volume.

Regulatory Picture

The regulatory environment has been permissive. NAIC issued model AI guidance in 2024 adopted in some form by approximately 35 states, requiring governance frameworks, decision documentation, human review for complex cases, and disparate impact avoidance — but not prohibiting AI claims processing or restricting its scope materially. A small number of class action lawsuits have challenged AI-driven settlements as bad faith, with mixed early results that have not slowed industry deployment.

State bad faith standards historically required showing unreasonable delay or denial. How these standards apply to algorithmic decisions is being tested in litigation that will substantially clarify through 2027-2028. The aggregate legal risk is manageable for the carriers.

What to Watch

Three signals over the next 90 days will indicate whether this trajectory continues:

  1. Q2 2026 personal auto carrier earnings call language. Watch for explicit touchless rate disclosures and comments on claims workforce trajectory.
  2. Vendor-disclosed customer count growth. Tractable, CCC, Hi Marley, and Snapsheet have been increasing the granularity of customer disclosure. Watch for accelerating customer acquisition.
  3. Claims adjuster job posting volume. Should continue declining from current levels. A material decline below 80% of 2024 baseline would confirm the trajectory.

The Larger Pattern

AI claims processing is not unique. It is the leading edge of broader back-office automation across the insurance value chain — underwriting, customer service, operations, billing — each on a similar but slightly delayed timeline. The aggregate impact on the roughly 2.8 million US insurance industry workforce (excluding agents) is substantial enough to be one of the major workforce transitions of the late 2020s.

The pattern repeats across regulated industries. The combination of structured input, structured output, clear unit economics, permissive regulation, and politically dispersed workforce produces the same displacement dynamic in medical coding, in customer service, and now in claims processing.

The displacement is happening quietly, on a CFO-driven timeline, without the press attention that the headline AI capability stories attract. The workforce impact is the largest part of the AI economic story that the public discussion is mostly missing.

Sources: