Legal Tech Briefing — 5 May 2026: California AI Ethics, Harvey Agents, Corgi AI Insurance

May 6, 2026
Legal Tech Briefing — 5 May 2026: California AI Ethics, Harvey Agents, Corgi AI Insurance

1. Top signals

  • California Bar proposes binding AI ethics rules across six Rules of Professional Conduct — first state to embed AI obligations in enforceable rules rather than guidance. Rule 1.1 requires lawyers to "independently review, verify, and exercise professional judgment regarding any output" used in client representation, with no carve-out for routine work. Rule 1.6 expands "reveal" to cover exposing client info to AI tools where retention or training creates material risk. Rule 3.3 explicitly extends the candor duty to AI-generated citations. Rule 5.1 requires firms to maintain functioning AI governance procedures. Comment period closed May 4 but rulemaking continues.
  • Moritz (formerly Arcline, YC-backed) raises $9M to operate as an AI-native NewMod law firm — pivoted from selling AI to traditional firms after concluding incentive structures blocked adoption. Investors include founders from Reddit, Dropbox, HuggingFace. Claims 100+ deals closed worth $2.3B aggregate, 4-hour average turnaround, $290M MSA drafted in 24h. 40+ "co-counselors" recruited from Fenwick, Cooley, Goodwin in three months. CEO Pamir Ehsas: "We want to be the biggest law firm in the world."
  • Harvey ships 500+ pre-built legal agents plus an Agent Builder (early access) — Winston Weinberg declares the industry "officially in the era of legal agents." Builder lets firms ground custom agents in their own knowledge and processes. Paired with a "Transformation Office" of legal-industry veterans for rollout management. Anique Drumright: agents "designed by lawyers who've done the work, not by prompt engineers."
  • Everlaw + Legora announce litigation workflow integration — surfaces Everlaw discovery documents directly inside Legora drafting and research, preserving user-permission governance across both platforms. Available to mutual customers in the coming months.
  • Corgi launches modular AI liability insurance ($108M raised, YC-backed) — endorsement covers algorithmic bias, AI hallucination/defamation, training-data misuse, data poisoning, autonomous-AI bodily injury, deepfake/synthetic media, AI IP, and regulatory defense. CEO Nico Laqua: "Where the liability is, we cover." Notes Orbital remains the only legal AI vendor offering insurance against its own outputs (residential property only).
  • Clio terminates 11+ year LawPay integration as of Aug 31, 2026 — fallout from 8am (LawPay's parent) acquiring Clio competitor MyCase in 2022. Clio Manage customers must migrate to Clio Payments or use LawPay outside the integration. Clio has been building out its own financial stack (Pay Later, Clio Capital).
  • Quentin Solt flags structural erosion at Thomson Reuters / RELX — research subscription revenue is dissolving as AI-assisted lawyers do less manual research; the strategic question for incumbents becomes participation in the verification layer of AI-native legal workflows. Fuller analysis promised this week.
Araik Tonian