On July 28, 2026, Crosby, a pioneering NewMod AI-native law firm, announced that it will take out professional liability insurance for its AI agents, enabling them to perform autonomous legal work without the current requirement that a human lawyer review every output. The announcement, reported by Artificial Lawyer, represents a watershed moment in the legal profession's relationship with artificial intelligence, effectively treating AI agents as autonomous legal professionals capable of delivering work product directly to clients, rather than as productivity tools that merely assist human attorneys. For personal injury law firms, the Crosby insurance model signals a fundamental shift in how the legal industry is reconceptualizing the division of labor between human attorneys and AI systems, and it raises urgent questions about malpractice exposure, ethical compliance, and the competitive landscape that will emerge as AI-native firms push the boundaries of autonomous practice.
Crosby's CEO, Ryan Daniels, framed the decision as the natural next chapter for a firm that has operated at the frontier of AI-legal integration since its founding. 'We are starting on the next chapter of Crosby: taking out professional liability insurance for our agents, so that they can do autonomous legal work,' Daniels stated. The rationale is straightforward: as AI agent capabilities have improved 'in leaps over the last few months,' the firm has concluded that requiring human lawyers to review every single work output is no longer necessary for the categories of 'complex but routine' matters that Crosby handles, including master service agreements, non-disclosure agreements, and partnership agreements. Daniels argues that if an AI agent is producing legal work that is sent directly to a client or integrated into a third-party platform, it is functionally indistinguishable from a human associate performing the same task, and should be insured as such.
The implementation of this insurance model is not straightforward. Crosby is currently working with bar associations, state regulators, auditors, and insurance providers to define the evaluation parameters that will determine whether an AI agent is eligible for professional liability coverage. These parameters will likely include benchmarks for accuracy, consistency, and compliance with jurisdictional rules of professional conduct, as well as protocols for handling edge cases, conflicting precedents, and client-specific risk factors that the AI may not have encountered in its training data. The firm is also developing what it calls the 'Multi-turn Negotiation Bench,' a standardized evaluation framework for measuring AI agent performance in contract negotiations, which could serve as an objective credentialing mechanism that insurers and regulators can reference when assessing whether an AI agent is competent to practice without human supervision.
Crosby's background provides context for the ambition of this initiative. The firm is backed by Sequoia Capital, Index Ventures, and Bain Capital Ventures, and has raised multiple rounds of funding to build a hybrid AI-law firm that combines proprietary AI agents with a thin layer of human attorneys for strategic oversight and client relationships. The firm operates on a fixed-fee model rather than traditional billable hours, arguing that AI-driven efficiency allows it to deliver high-speed, cost-effective legal services at price points that traditional law firms cannot match. Crosby has also launched 'Crosby Intelligence,' a research group focused on agentic attorneys, and has been an active contributor to industry benchmarking efforts, including the 'Legal Agent Bench' that evaluates AI performance across standardized legal tasks.
For personal injury law firm leadership, the Crosby insurance model carries three layers of strategic significance. First, the decision to insure AI agents as legal professionals directly challenges the prevailing assumption that AI is a tool for human attorneys rather than a substitute for human judgment, and it suggests that the malpractice insurance industry will soon need to develop actuarial models for AI-driven legal work. PI firms should monitor whether their own malpractice carriers begin to ask about AI usage in underwriting questionnaires, and whether premiums or coverage terms will be adjusted based on the extent to which a firm relies on AI-generated work product. Second, the Crosby model raises important ethical questions about unauthorized practice of law and the duty of competent representation. If an AI agent can be insured for legal liability, it may also be subject to bar discipline, and PI firms should track whether state bar associations issue guidance on whether AI agents performing legal work without human supervision constitute a violation of rules that require attorneys to provide competent representation and maintain direct supervision over non-lawyer assistants. Third, the competitive implications are profound. If Crosby and other AI-native firms can deliver routine legal work at a fraction of traditional cost while maintaining professional liability coverage, the economic pressure on traditional law firms, including PI firms, will intensify. PI firms that have invested in AI for document review and case management should evaluate whether their current tools are capable of the autonomous output quality that Crosby is achieving, and whether the firm's own malpractice coverage would extend to AI-generated work product that is sent to clients without human review. The Crosby insurance model is not merely an operational experiment; it is a declaration that AI agents have arrived as autonomous legal professionals, and the profession must now adapt its ethical, insurance, and regulatory frameworks to accommodate this new reality.
