GC AI Has Raised Nearly $72M to Build AI for In-House Legal Teams

GC AI, the startup founded by former Replit general counsel Cecilia Ziniti and engineer Bardia Pourvakil, has raised nearly $72 million across three rounds. Its latest, a $60 million Series B co-led by Scale Venture Partners and Northzone, valued the company at $555 million. The company is building AI software for corporate legal teams, where trustworthy use of confidential documents is as central to the sale as a model’s ability to generate an answer.

The company says it serves roughly 2,100 companies, up from about 900 a year earlier, and has grown 400% year over year. Its named customers include Lockheed Martin, Time, Eventbrite, Vercel and Gusto. Those are company-supplied growth and customer figures, not independently audited operating results.

A product brief shaped by an in-house lawyer

Ziniti left Replit on November 1, 2023, and incorporated San Mateo-based GC AI about a week later. She had worked as an in-house lawyer before founding the company, while Pourvakil was an engineer she had worked with at Replit. Their split maps onto the company’s premise: one founder knew the daily legal workflow; the other built the software.

GC AI began with an AI assistant for corporate legal departments, then added contract analysis and a product for handling requests sent to legal teams. Ziniti’s early experience teaching lawyers to use ChatGPT helped define the intended difference from a general-purpose chatbot: answers needed accurate citations, document-level backing and a professional legal tone.

Trust is the commercial constraint

Legal departments hold confidential contracts, internal requests and other sensitive company material. GC AI says it pursued SOC 2 compliance early, uses data-isolation protections and does not train its models on customers’ confidential information. Those are the company’s stated safeguards; the account does not provide independent testing or security-audit results for them.

Where GC AI draws its product boundary

• Its contract-intelligence product is designed to analyze a company’s own documents.
• Its legal-request product is meant to manage and respond to requests submitted to corporate legal departments.
• Nearly 30% of GC AI seats are used outside legal, including by HR and finance executives, according to Ziniti.

A different buyer from law-firm AI

Ziniti positions GC AI against legal-AI companies Harvey and Legora by customer type rather than just model capability. Harvey and Legora initially focused on law firms, while GC AI has concentrated on corporations and their in-house teams. That distinction affects the data problem: GC AI’s contract product is meant to work within one company’s records, whereas law-firm tools may need to operate across materials from thousands of clients.

The financing gives GC AI resources to pursue that narrow workflow focus at a larger scale. It does not settle the harder product question: whether citations, isolated data and legal-specific tone will prove sufficient for corporate teams to rely on AI across consequential legal work. The company’s reported customer growth suggests demand, but the evidence offers no independent measure of answer accuracy or of how much legal work customers ultimately hand to the system.