Harvey, a startup building an AI-powered “co-pilot” for lawyers, has raised $100 million in a Series C round led by Google's corporate ventures arm GV.
The round also included participation from leading angel and venture capital firms, including OpenAI, Kleiner Perkins, Sequoia Capital, Elad Gil and SV Angel, bringing Harvey's total funding to $206 million and valuing the company at $1.5 billion.
In a Tuesday post on Harvey's official blog, co-founders Winston Weinberg and Gabriel Pereira said most of the new funding will be used to grow Harvey's headcount and expand paid services to new geographies, as well as collect and curate data to build and train “domain-specific” AI models.
“This investment will enable Harvey to continue to expand and improve our AI-powered technology across business functions and regions,” Weinberg and Pereyra said. “We will use this new funding to invest in the engineering, data and domain expertise essential to building AI-native systems that facilitate the most complex knowledge work. We will also deepen partnerships with both cloud providers and model providers to integrate additional models into Harvey and expand training collaborations to continually improve the effectiveness of our models.”
Weinberg, a former securities and antitrust lawyer at law firm O'Melveny & Myers, and Pereira, a former research scientist at DeepMind, Google Brain (another AI group at Google) and MetaAI, founded San Francisco-based Harvey in 2022. Weinberg and Pereira were roommates when Pereira introduced Weinberg to OpenAI's GPT-3 text-generation system, which Weinberg realized could be used to improve legal workflows.
Powered by OpenAI's GPT-4 family of models, Harvey can answer legal questions phrased in natural language, such as, “What is the difference between an employee and an independent contractor under the Fourth Circuit?” or “Does this clause in my lease violate California law? If so, please rewrite it so that it does not violate the law.”
Harvey also provides tools to automatically extract information from court records, automatically search for legal documents to support arguments in court, and generate first drafts of filings that incorporate information and citations from legal databases.
While it's a powerful tool in theory, it also comes with risks. Due to the sensitive nature of most legal disputes, some lawyers and law firms may be reluctant to give a tool like Harvey access to legal documents. There's also the issue that language models tend to spit out harmful content or fabricated facts, which would be particularly unjust in court and could lead to perjury charges.
That's why Harvey comes with a disclaimer: This tool is not intended to provide legal advice to non-lawyers and should be used under the supervision of a licensed attorney.
Harvey has competitors, too. Casetext uses AI, primarily OpenAI models, to find legal cases and help with general legal research tasks and drafting briefs. More specialized tools like Klarity use AI to take the tedium out of contract review. At one point, startup Augrented was exploring how it could use OpenAI models to summarize legal notices and other sources of information in plain English to help tenants protect their rights.
But Weinberg and Pereira claim Harvey is growing well, with tens of thousands of lawyers using it “daily” at law firms and consultancies including Allen & Overy, McFarlane, Ashurst, CMS, Reed Smith and PwC. Annual recurring revenue has tripled since December, the co-founders said in a blog post, and Harvey's headcount has tripled.
In early June, The Information reported that Harvey was looking to raise $600 million at a valuation of “at least” $2 billion, partly to acquire a legal research service called vLex to train its AI products, but that plan fell through and the Series C was significantly reduced.