The Intelligence Asymmetry: Social Inflation Is a Misdiagnosis
The Real Driver Is a Plaintiff-Side AI Ecosystem the Defence Industry Has Not Matched
EvenUp, a generative AI platform designed for personal injury firms, achieved a $2 billion valuation in 2025. This single metric captures a structural shift in American litigation. The plaintiff bar has spent billions on technology to aggregate settlement data, standardise demand packages and model carrier-specific risk tolerances. The defence side remains fragmented. Individual insurance carriers and their panel counsel operate within data silos that obscure broader patterns in judicial behaviour and plaintiff strategy.1
This divergence is driven by asymmetric AI adoption and the ways both sides collect, aggregate and utilise litigation data.1 The insurance industry has long attributed rising liability costs to “social inflation” and “nuclear verdicts” as external, uncontrollable phenomena driven by shifting societal norms and anti-corporate bias. Emerging evidence suggests a different explanation: much of what the market experiences as uncontrolled cost growth is the direct result of coordinated technological advantage on the plaintiff side.
Plaintiff firms have transitioned from isolated practices to a technologically integrated ecosystem. They leverage AI to standardise aggressive demand packages and use crowdsourced settlement data to identify carrier-specific leverage points. The defence side has not matched this coordination. Individual insurance carriers and their counsel operate in isolation, obscuring broader market patterns and failing to develop shared intelligence about plaintiff tactics or judicial outcomes.1
This structural imbalance drives up litigation costs and sends inconsistent signals to the market. When defence outcomes vary widely across materially similar cases, the plaintiff bar identifies these variations as opportunities. High settlements (even outliers) become the new reference points for future negotiations. The market amplifies what should be treated as noise into trend.
The Plaintiff-Side Technological Vanguard: Coordination at Scale
The plaintiff bar’s adoption of artificial intelligence is rooted in economic incentive. Operating on a contingency fee basis, plaintiff firms maximise returns by adopting any technology that reduces manual labour, shortens case lifecycle and increases settlement values.4,5 By 2024, approximately 30% of legal professionals reported using AI technology, a 19% increase from the previous year, with personal injury practices at the forefront of this adoption curve.5
Automated Demand Engineering
The demand letter, historically one of the most time-consuming tasks in personal injury practice, has become automated. Platforms such as EvenUp, Supio and Novo ingest thousands of pages of unstructured data (medical records, police reports, billing statements) and generate settlement-ready demand packages in minutes.6,8,9 These tools extract the most salient facts, map injuries to specific ICD-10 codes, and highlight pain and suffering narratives that manual review may overlook.8,10
EvenUp’s “Mirror Mode” allows firms to train AI to draft documents that match their distinctive tone and style. The firm maintains efficiency without sacrificing brand consistency.3 These systems proactively identify missing medical records or treatment gaps, allowing firms to maximise claim value before the initial demand. This level of preparation forces adjusters into a defensive posture, presented with data-rich demands that are difficult to challenge without equivalent technological support.13
Crowdsourcing Private Settlements: Neutralising Information Asymmetry
The most notable strategic shift in the plaintiff bar is the move away from negotiating cases in isolation.1 Historically, the defence side held an information advantage. Insurers handled thousands of claims and understood the “going rate” for specific injuries in specific venues. Plaintiff-side AI firms have neutralised this advantage by convincing thousands of firms to share their actual, anonymised settlement data.3
This crowdsourced data allows AI platforms to analyse not just public jury verdicts (which represent a tiny fraction of civil cases) but the private settlements that constitute the vast majority of resolutions.3,14 By synthesising this collective intelligence, plaintiff attorneys can identify venue-specific leverage points, judicial tendencies and carrier-specific settlement patterns.1 They can determine which adjusters are most likely to settle early and at what values, allowing them to calibrate their demands accordingly.1
This coordination enables a standardisation of aggressive positions across large inventories of cases. No potential value remains on the table. The scale of investment supporting this shift is substantial. EvenUp’s $2 billion valuation signals that the plaintiff bar has reached a tipping point in its ability to leverage data as a strategic asset.3
The Defence Intelligence Gap: Fragmentation and Operational Friction
The insurance defence industry stands in stark contrast to the coordinated ecosystem of the plaintiff bar. This fragmentation is both organisational and industry-wide, creating systemic weaknesses that prevent the defence from countering the plaintiff side’s data-driven strategies.1
The Internal Silo Problem
Within a typical insurance organisation, data is siloed between departments and individual adjusters. A claims professional in one region may have little visibility into how similar cases are being resolved in another, or even how different adjusters within the same office are valuing identical injuries.1 This lack of internal communication produces uncertain valuation anchors. Negotiators struggle to distinguish between a fair evolution in case value and inflated “drift” caused by aggressive plaintiff tactics.1 When defence outcomes vary widely across materially similar cases, it creates inconsistent market signals.1
The plaintiff bar, using its coordinated AI tools, identifies these variations as opportunities. High settlements quickly become new reference points for future negotiations.1 Because the defence side lacks a shared view of its own portfolio’s value, it often fails to recognise when a specific strategy is being tested across multiple jurisdictions until it has already become established as a trend.1
The Hourly Billing Barrier to AI Adoption
Defence law firms traditionally operate on an hourly billing model. This creates a misalignment of incentives. Time-saving technologies that would benefit the insurer by reducing legal spend may be perceived as a threat to the firm’s revenue.4 This friction leads to a cautious stance toward AI adoption.
Whilst plaintiff firms are integrating generative AI to summarise medical records and draft pleadings, many defence teams are still manually sifting through voluminous discovery materials.4,15,16 The manual process is slower and more prone to human error. AI can identify pre-existing conditions or inconsistencies in a plaintiff’s medical history in seconds.15,16
Deconstructing Social Inflation as a Negotiated Phenomenon
The insurance industry has long used the term “social inflation” to describe the rise in liability claims costs above general economic inflation.18,19 This label encompasses several factors, including anti-corporate bias and the rollback of tort reform. The available evidence suggests that much of what is attributed to social inflation is instead a portfolio phenomenon driven by the intelligence gap.1
Quantifying Loss Severity
Actuarial research by the Insurance Information Institute and the Casualty Actuarial Society indicates that “legal system abuse” (‘LSA’) contributed between $231.6 billion and $281.2 billion in increased liability insurance losses between 2014 and 2023.20 This surge outpaces general economic inflation as measured by the CPI-U.20
The data confirms that these losses are driven by claim severity rather than frequency.20,21 The number of claims filed has remained relatively stable or declined in some segments. The average payout per claim has soared.20,21 This increase in severity results from more effective anchoring strategies employed by the plaintiff bar, often supported by AI-generated demand packages that highlight the most extreme potential damages.10,22
Methodological Concerns with the LSA Framework
The insurance industry has attributed escalating liability costs to legal system abuse, pointing to sophisticated plaintiff tactics, third-party litigation funding and technology-driven demand packages as culprits behind claim severity that outpaces economic inflation.20
This explanation warrants scrutiny. The LSA framework suffers from procedural weaknesses beginning with definitional circularity. Unexplained loss increases are simply labelled as “abuse” without rigorous proof of causation. The research fails to control adequately for legitimate cost drivers: medical inflation (which has consistently outpaced CPI for decades), escalating vehicle repair costs for technology-laden modern automobiles, genuine increases in injury severity, and wage growth affecting lost earnings calculations. Triple-I is an industry advocacy organisation rather than an independent research body. Its framing of increased costs as “abuse” may reflect an ideological perspective rather than objective analysis.
An alternative explanation for rising insurance premiums challenges the legal system abuse narrative. Cost escalation may result instead from systemic inefficiencies and poor coordination within the defence bar and insurance industry itself. Whilst plaintiff attorneys have invested heavily in technology, data analytics and coordinated litigation strategies, the defence side has remained fragmented. Individual insurers and their counsel work in silos without sharing intelligence or best practices. This lack of strategic coordination means defence counsel repeatedly reinvent solutions for similar cases, fail to leverage collective data on settlement patterns, and miss opportunities for early intervention that would reduce claim severity.
Many insurers have been slow to adopt the technological tools now standard among plaintiff firms, relying instead on outdated case management systems and reactive rather than proactive defence strategies. The result is not legal system abuse but a predictable market response: when one side invests in efficiency and intelligence whilst the other does not, costs naturally rise. The premium increases may reflect the insurance industry’s own failure to modernise and coordinate its defence efforts, rather than any distortion of the legal system.
The Role of Third-Party Litigation Funding
The escalation of litigation costs is further fuelled by the growth of third-party litigation funding, a global industry that enables investors to finance lawsuits in exchange for a portion of the settlement.18,23,24 TPLF provides the capital necessary for plaintiff firms to invest in the technology widening the intelligence gap.1,24 Analysis by the National Insurance Crime Bureau suggests that 75% of assessed insurance companies have been directly targeted by litigation marketing campaigns backed by funders.24
Some observers suggest this funding transforms individual personal injury claims into an asset class, incentivising longer litigation cycles and higher settlement demands as funders seek to maximise their return on investment.18,23
Proponents of third-party litigation funding argue that it serves a vital access-to-justice function, particularly for claimants lacking the financial resources to pursue meritorious claims against well-funded defendants. Without such funding mechanisms, many legitimate claims would remain unpursued due to prohibitive litigation costs, effectively denying redress to those who have suffered genuine harm. TPLF helps level the playing field, enabling individuals and smaller entities to assert their legal rights against corporations and insurers with substantially greater financial capacity. From this perspective, funding democratises access to the legal system rather than distorting it, ensuring that the ability to seek justice is not determined solely by financial means.
Towards a Defence AI Ecosystem: Restoring the Strategic Balance
To counter the coordinated intelligence of the plaintiff bar, insurance organisations and defence firms must adopt AI-driven tools and foster industry-wide data sharing.1 This transition requires moving beyond AI as a tool for document review. Integration into the core of negotiation and portfolio management is essential.1
Predictive Modelling and Valuation Consistency
Emerging platforms designed for the defence side focus on restoring valuation anchors and providing defensible settlement recommendations.13,25 Systems like SigmaSight provide AI-driven analytics that generate risk and valuation ranges grounded in economic data, venue trends and specific injury types.13 These tools allow claims professionals to negotiate with confidence. Every offer is supported with a structured Offer Package that clearly explains the rationale to mediators and opposing counsel.13
AI for predictive settlement modelling enables insurers to estimate settlement ranges and probabilities more accurately, tightening reserve bands and reducing late-stage financial surprises.25 By identifying outlier cases early, these systems allow defence teams to anticipate plaintiff strategies before they result in large verdicts.13
Counsel Collaboration and Pattern Recognition
AI can serve as a bridge between insurance carriers and their defence counsel. By sharing AI-backed insights, claims professionals can ensure their attorneys are aligned on valuation and strategy from the outset.13 This collaboration allows the defence to develop more effective responses against specifically aggressive plaintiff firms.26 CLARA Analytics provides detailed scorecards for defence counsel, benchmarking them against industry peers and analysing their past performance against specific plaintiff attorneys.26
This shift addresses the pattern recognition deficit on the defence side.1 By aggregating data across portfolios, AI can identify when a particular plaintiff firm is deploying a new tactical narrative (such as “junk science” theories or novel pain and suffering arguments), allowing the defence to develop a coordinated response rather than reacting to each case in isolation.1
Ethical Guardrails and the Judicial Response to AI
The rapid integration of AI into legal practice has created new ethical risks, with AI hallucinations emerging as the most visible: generative tools fabricate non-existent case citations.28,29,30 High-profile incidents like Mata v. Avianca and Johnson v. Dunn resulted in substantial judicial sanctions and a sharp rebuke of attorneys who failed to exercise appropriate oversight.29,30
Human-in-the-Loop Oversight
These judicial developments highlight a requirement for AI adoption on both sides: the human-in-the-loop model.27,31 Ethical guidelines from the American Bar Association and state bars emphasise that AI tools must remain under human authority, oversight, and control.5,32 For the defence side, where professional credibility and strict confidentiality are paramount, this human oversight is non-negotiable.16,17,27 AI should perform data-intensive tasks of extraction and summarisation. Final legal judgment and strategic decisions must rest with the attorney.4,16
Privilege and the Digital Stranger
A concern for insurers is the potential waiver of attorney-client privilege when using public AI platforms.17,33 Public models often retain uploaded data for training purposes, potentially exposing sensitive litigation strategies to a “digital stranger” outside the protected relationship.17 To mitigate this risk, defence organisations are moving toward private or firm-built AI systems that operate under strict contractual relationships ensuring data segregation and confidentiality.17,34 A private AI system functions more like an agent of the firm (an expert witness or e-discovery consultant) rather than an unregulated third party.17
Regulatory Evolution and the 2025 Roadmap
As the impact of AI on litigation becomes clearer, regulators are taking proactive steps to ensure market stability and consumer protection.35,36 The National Association of Insurance Commissioners (‘NAIC’) has identified AI governance as a top priority in its 2025 roadmap, “Securing Tomorrow: Advancing State-Based Regulation”.35
NAIC Initiatives
The NAIC’s 2025 strategy focuses on enhancing financial governance and modernising risk-based capital frameworks to address the demands of the evolving market.35 Specifically, the NAIC is working on frameworks for third-party data and predictive models, emphasising outcome transparency.36 States such as New York, Colorado and Connecticut are expected to lead in “AI outcomes testing,” requiring insurers to demonstrate that their use of AI does not result in unfair bias or discriminatory practices.36
The NAIC is expected to introduce a new privacy protections model law by late 2025, addressing the risks associated with the massive amounts of data (estimated to reach 181 zettabytes annually) being generated and consumed by AI systems.36,38 This regulatory pressure will force insurers to be more deliberate and transparent in their use of AI for claims and litigation management.36,39
Re-Examining Collective Intelligence: The Industry Data Mandate
The final requirement for restoring balance is the need for industry-wide coordination on the defence side.1 If the plaintiff bar is benefiting from collective intelligence through shared settlement repositories, the defence industry must confront whether its traditional norms of isolation are inadvertently causing financial harm.1
Other sectors of the insurance industry already rely on pooled data to manage risk. The National Insurance Crime Bureau operates shared fraud detection systems. Catastrophic modelling relies on shared databases.1 Litigation negotiation remains a notable exception, where each carrier negotiates without the context of the broader market.1
Establishing contributory databases or shared benchmarking would allow the defence side to achieve parity with the tools the plaintiff bar is already using.1 This does not require perfect data, but rather a commitment to coherence and proportionality.1 By creating a unified front of data, insurers can prevent the anchoring of settlements at unsustainable levels and ensure that individual outliers do not recalibrate the entire market’s expectations.1
The ultimate goal of adopting AI and data-sharing is to strengthen, rather than replace, professional judgment.1,13,40 AI functions as a navigational tool for adjusters, providing structured insights that help them analyse complex files more effectively.25,26 By automating the repetitive aspects of case analysis (such as medical billing review and chronology generation), AI frees up human professionals to focus on high-value tasks of strategy, advocacy, and negotiation.13,41,42
Conclusion: Restoring Equilibrium in the AI Era
The growing intelligence gap in litigation is not an inevitable byproduct of technological advancement. It is the result of a coordinated strategic shift by the plaintiff bar that has yet to be fully met by the defence.1 Social inflation is a real and measurable force. Its impact, however, is amplified by the defence side’s internal fragmentation and lack of data-driven negotiation tactics.1
Restoring equilibrium requires a dual response. The defence must integrate AI-driven tools into the claims and litigation workflow to improve valuation consistency, reduce operational costs, and sharpen defence arguments.13,15 Beyond the firm level, the insurance industry must abandon the silo mentality and develop mechanisms for collective intelligence and shared benchmarking to counter the plaintiff bar’s coordinated advantage.1
By adopting a proactive approach to AI and data sharing, the defence side can transition from a reactive posture (often blamed on external societal shifts) to one of professional control and financial stability.1,13 The future of casualty litigation will be defined by those who can most effectively synthesise the vast oceans of available data into actionable, defensible intelligence.26,38 Restoring balance at the negotiating table is not merely a technological challenge. It is a structural mandate for the preservation of a fair and sustainable legal system.1,40
References
1 The Defense Intelligence Gap: Why Litigation Portfolio Values Are Accelerating in an AI Era, https://riskandinsurance.com/the-defense-intelligence-gap-why-litigation-portfolio-values-are-accelerating-in-an-ai-era/
2 Turning the tide on social inflation and rising liability insurance costs—Lockton, https://global.lockton.com/us/en/news-insights/turning-the-tide-on-social-inflation-and-rising-liability-insurance-costs
3 From Startup To $2 Billion: EvenUp Is Transforming Personal Injury, https://www.techlawcrossroads.com/2025/10/from-startup-to-2-billion-evenup-is-transforming-personal-injury-practice/
4 How Plaintiff Law Firms Are Outpacing Defense Firms in AI Adoption—And What to Do About It, Wisedocs, https://www.wisedocs.ai/blogs/how-plaintiff-law-firms-are-outpacing-defense-firms-in-ai-adoption
5 Using emerging AI tech to prompt your future, Advocate Magazine, https://www.advocatemagazine.com/article/2025-august/using-emerging-ai-tech-to-prompt-your-future
6 Eve | Legal AI for plaintiff firms, https://www.eve.legal/
7 The Top 5 Benefits of AI for Personal Injury Law Firms, Eve Legal, https://www.eve.legal/blogs/the-top-5-benefits-of-ai-for-personal-injury-law-firms
8 AI Demand Letter Drafting Platform for Personal Injury Firms, Supio’s AI, https://www.supio.com/products/ai-demand-letter
9 AI Demand Letter Tools and Strategies for Personal Injury Firms, CasePeer, https://www.casepeer.com/blog/ai-demand-letter/
10 EvenUp | Personal Injury AI — From Intake to Resolution, https://www.evenuplaw.com/
11 Negotiation Preparation™ | AI-led Settlement Negotiations Platform, EvenUp, https://www.evenuplaw.com/products/negotiation-preparation
12 Improve Demand Letter Effectiveness with AI-Powered Insights, precedent.com, https://precedent.com/the-rise-of-pi-automation-2025-outlook/
13 Claim Professionals, SigmaSight, https://www.sigmasight.ai/claim-professionals
14 Policyholders Are Not to Blame for “Social Inflation”, https://www.pillsburylaw.com/en/news-and-insights/policyholders-not-to-blame-for-social-inflation.html
15 AI for Insurance Carriers | CaseMark—Pre-Approved AI for Defense Teams, https://www.casemark.com/solutions/insurance-carriers
16 Enhancing Legal Defense Strategies with AI Medical Summaries, Wisedocs, https://www.wisedocs.ai/blogs/enhancing-legal-defense-strategies-with-ai-medical-summaries
17 Digital Strangers in Litigation: Does Sharing with AI Breach Privilege?, TransPerfect Legal, https://www.jdsupra.com/legalnews/digital-strangers-in-litigation-does-6854504/
18 Insurance Topics | Social Inflation, NAIC, https://content.naic.org/insurance-topics/social-inflation
19 Social Inflation: Navigating the evolving claims landscape, The Geneva Association, https://www.genevaassociation.org/sites/default/files/social_inflation_brief_web.pdf
20 Legal System Abuse, Not Just Economic Inflation, Drives Liability Insurance Losses by More than $230 Billion Over Past 10 Years, Insurance Information Institute, https://www.iii.org/press-release/legal-system-abuse-not-just-economic-inflation-drives-liability-insurance-losses-by-more-than-230-billion-over-past-10-years-new-triple-i-casualty-actuary-society-analysis-shows-103025
21 What Is the Evidence for Social Inflation?, RAND, https://www.rand.org/pubs/research_reports/RRA2645-1.html
22 Three Key Drivers of Social Inflation, AJG United States, https://www.ajg.com/news-and-insights/features/social-inflation-nuclear-verdicts-drivers/
23 4 Social Inflation Drivers Contributing to Rising Claim Costs, Travelers Insurance, https://www.travelers.com/resources/business-topics/insuring/4-factors-causing-social-inflation
24 NICB Warns Insurance Industry of Fraud Facilitated by Third-Party Litigation Funding, https://www.nicb.org/news/news-releases/nicb-warns-insurance-industry-fraud-facilitated-third-party-litigation-funding
25 Negotiated Settlement Impact AI Agent for Claims Economics in Insurance, insurnest, https://insurnest.com/agent-details/insurance/claims-economics/negotiated-settlement-impact-ai-agent-for-claims-economics-in-insurance/
26 Litigation Analytics – Turning Data into a Competitive Advantage for Casualty Insurers, Genre, https://www.genre.com/us/knowledge/publications/2025/august/litigation-analytics-turning-data-into-a-competitive-advantage-for-casualty-insurers-en
27 AI, Trust, and the Human Touch: Wisedocs’ Webinar with CLM, https://www.wisedocs.ai/blogs/ai-trust-and-the-human-touch-clm-webinar
28 Artificial Intelligence or innocent ignorance? Hard lessons yield best practices, Clark Hill, https://www.clarkhill.com/news-events/news/artificial-intelligence-or-innocent-ignorance-hard-lessons-yield-best-practices/
29 Will AI Render Lawyers Obsolete?, New York State Bar Association, https://nysba.org/will-ai-render-lawyers-obsolete/
30 Federal Court Turns Up the Heat on Attorneys Using ChatGPT for Research, https://www.esquiresolutions.com/federal-court-turns-up-the-heat-on-attorneys-using-chatgpt-for-research/
31 How AI Is Reshaping Litigation: Insights from DRI’s Senior Living Seminar, Excelas, https://excelas1.com/how-ai-is-reshaping-litigation-insights-from-dris-senior-living-seminar/
32 Law-Related Artificial Intelligence: Ethics Issues, McGuireWoods, https://media.mcguirewoods.com/publications/Ethics-Programs/88951932.pdf
33 Your AI-Powered Practice: Trial Preparation, Ethics, and the Future, Sam Aguiar Injury Lawyers, https://aguiarinjurylawyers.com/your-ai-powered-practice-trial-preparation-ethics-and-the-future/
34 The AI-Enabled Insurance Defense Firm: A Strategic Roadmap for an Era of Disruption, https://www.performlaw.com/law-firm-best-practices-blog/the-ai-enabled-insurance-defense-firm-a-strategic-roadmap-for-an-era-of-disruption
35 NAIC Announces 2025 Initiatives, https://content.naic.org/article/naic-announces-2025-initiatives
36 2025 Insurance Regulatory Outlook, Deloitte US, https://www.deloitte.com/us/en/services/consulting/articles/insurance-regulatory-outlook.html
37 Commissioner Lara Announces Major Reforms to the Intervenor Process, California Department of Insurance, https://www.insurance.ca.gov/0400-news/0100-press-releases/2025/release059-2025.cfm
38 The Two Faces of AI/Articles/CLM Magazine, Claims and Litigation Management Alliance, https://www.theclm.org/Magazine/articles/the-two-faces-of-ai/3218
39 Overview | Data Protection & AI, Lewis Rice, https://www.lewisrice.com/data-protection-ai
40 DRI The Voice, August 2025, https://www.dri.org/newsletters/the-voice/2025/august
41 AI-Powered Defense (Part 2): Better Outcomes, Less Effort, More Profit, DRI Communities, https://community.dri.org/events/event-description?CalendarEventKey=98853280-3b56-4cb7-8de4-01996551889b&Home=%2Fevents%2Fcalendar
42 The Rise of AI/Articles/CLM Magazine, Claims and Litigation Management Alliance, https://www.theclm.org/Magazine/articles/the-rise-of-ai-in-workers-comp-claims/2873