Recommended Articles
Tech & AI

Legal AI Goes Firmwide: Harvey, Legora, Blackletter — and California's New Rules

By Angeled · November 10, 2026 · Tech & AI
Legal AI Goes Firmwide: Harvey, Legora, Blackletter — and California's New Rules
Legal AI Goes Firmwide: Harvey, Legora, Blackletter — and California's New Rules

What happened

Legal AI crossed a threshold in November 2026: it stopped being a pilot project and became firmwide infrastructure at major law firms.

UK firm Addleshaw Goddard will deploy legal AI Harvey firmwide to around 3,400 lawyers and business support staff starting in January 2027, and will co-develop features with Harvey. On the same day, Fieldfisher adopted competitor Legora firmwide. Two large firms, two competing platforms, the same decision on the same day: the whole firm, not a team.

Separately, Blackletter — a New York law firm launched in 2026, staffed by attorneys and built around its proprietary BlackletterOS AI, with fixed fees — raised a $6 million seed round led by Lerer Hippeau, reported on November 9, 2026.

And the rules are catching up. California SB 574, signed on September 30 by Governor Newsom and effective January 1, 2027, bars delegating law practice to AI, requires lawyers to verify AI outputs, requires them to personally verify citations, requires disclosure of AI use to courts, and bars entering confidential information into unrestricted systems.

Regulators are also moving. The SEC's FY2027 exam priorities pivot to a threat-based AI focus, and the Treasury's Financial Services AI Risk Management Framework contains 230 control objectives.

What firmwide deployment means

There is a large difference between a pilot and a firmwide deployment. A pilot involves a small group of enthusiastic users, a limited set of tasks, and an easy exit if it does not work. A firmwide deployment to around 3,400 people involves training, support, security review, workflow redesign and a commitment that the tool will be part of how the firm works every day.

Addleshaw Goddard's decision to co-develop features with Harvey adds another layer. The firm is not only buying a product; it is helping to shape it. That suggests the firm expects the platform to be a long-term part of its practice, and that it wants the product to fit its specific workflows rather than the other way around.

Blackletter: a firm built around AI

Blackletter is a different model. It is a law firm launched in 2026, not a technology vendor selling to firms. Its attorneys work with BlackletterOS, its proprietary AI system, and the firm charges fixed fees.

For founders, Blackletter is a case study in business-model design. The technology alone is not the moat. The combination of technology, pricing and firm structure is what makes the model distinct.

California SB 574: the guardrails

California SB 574, effective January 1, 2027, sets out rules that every lawyer using AI in California will need to follow. The law bars delegating law practice to AI — the lawyer remains responsible. It requires lawyers to verify AI outputs, and to personally verify citations. It requires disclosure of AI use to courts. And it bars entering confidential information into unrestricted systems.

The regulatory backdrop

The SEC's FY2027 exam priorities, with their pivot to a threat-based AI focus, and the Treasury's Financial Services AI Risk Management Framework, with 230 control objectives, show that legal AI is not being regulated in isolation. Financial regulators are also building frameworks for how AI is used, governed and examined in their sectors.

What this means for founders

For legal-tech founders, firmwide deployments are the strongest possible validation — and the hardest to win. A firm that deploys to 3,400 people has done extensive diligence on security, accuracy and workflow fit. Founders pitching firms should expect that level of scrutiny, and should build for it from the start: audit trails, citation verification, confidentiality controls and clear disclosure of when AI is used.

For founders outside legal tech, the lesson is about the adoption curve. A technology that was a pilot last year can become firmwide infrastructure this year. The window to establish a position in a professional-services market can close quickly once the large firms choose their platforms.

What this means for investors

The Harvey and Legora deployments suggest that legal AI has reached the stage where revenue is driven by firmwide contracts, not individual seats. That changes the unit economics: larger deals, longer commitments, and deeper integration with the customer's workflows.

Blackletter's seed round points to a second investment thesis: AI-native firms that rebuild the business model, not only the tooling. Investors should ask whether a legal-tech company is selling software to firms, or building a firm itself — and which model has the stronger moat.

Regulatory risk is also part of the analysis. SB 574 and the SEC and Treasury frameworks do not ban legal AI, but they set conditions on its use. Companies that build verification, disclosure and confidentiality into their products will be better positioned than those that treat compliance as an afterthought.

Angeled.com connects founders with investors.

kimkj.com briefing, November 9–10, 2026
Business Wire, via crosswalk.to, November 9–10, 2026
NC Lawyers Weekly, November 9–10, 2026
briefglance.com, November 9–10, 2026

Read Next