In a Series C investment headed by GV, Google’s corporate venturing arm, Harvey, a firm creating an AI-powered “copilot” for attorneys, has raised $100 million.
With the participation of aggressive angel investors and venture capital firms OpenAI, Kleiner Perkins, Sequoia Capital, Elad Gil, and SV Angel, the tranche increases Harvey’s overall fundraising to $206 million and values the company at $1.5 billion.
Co-founders Winston Weinberg and Gabriel Pereyra stated in a post on Harvey’s official blog on Tuesday that the majority of the additional funding will go toward hiring more staff members and extending the company’s paid services to new regions, all while gathering and analyzing data to create and train “domain-specific” AI models.
“This investment will enable Harvey to continue scaling and improving our AI-powered technology across business functions and geographies,” Weinberg and Pereyra said. “We will use this new capital to invest in the engineering, data and domain expertise that are fundamental to building AI-native systems that facilitate the most complex knowledge work. We will also deepen our partnerships with both cloud and model providers to integrate additional models into Harvey and broaden our training collaborations to continue improving model efficacy.”
Harvey was founded in San Francisco in 2022 by Weinberg, a former securities and antitrust attorney at the law firm O’Melveny & Myers, and Pereyra, a former research scientist at DeepMind, Google Brain (another of Google’s AI departments), and Meta AI. Weinberg and Pereyra share a room; after Pereyra demonstrated OpenAI’s GPT-3 text-generating system to Weinberg, the latter recognized its potential for streamlining legal operations.
With the support of OpenAI’s GPT-4 model family, Harvey is able to respond to legal inquiries expressed in natural language, such as “Tell me what the differences are between an employee and independent contractor in the Fourth Circuit” and “Tell me if this clause in a lease is in violation of California law, and if so, rewrite it so it is no longer in violation.” Harvey provides tools that can also automatically locate papers that support court arguments, extract material from trial transcripts, and create preliminary file drafts that include information and citations from legal databases.
Theoretically, it’s powerful stuff. However, it’s also risky. Some attorneys and law firms might be hesitant to provide any case papers to a tool like Harvey because most legal conflicts are delicate. The tendency of language models to propagate harmful and fabricated information is another issue. These statements would be especially offensive, if not harmful, in a court of law.
Harvey comes with a disclaimer because of this: the tool should only be used under the guidance of licensed attorneys and isn’t intended to give legal advice to non-lawyers.
Harvey has some opposition. Casetext finds legal cases and helps with common legal research duties and brief preparation by utilizing AI, mostly OpenAI models. AI is being used in more surgical tools like Klarity to remove tedious contract review processes. In an effort to support apartment tenants in asserting their legal rights, company Augrented previously investigated the possibility of using OpenAI models to translate court notices and other sources into understandable English.
However, Weinberg and Pereyra assert that tens of thousands of attorneys at various legal firms and consultancies, such as Allen & Overy, Macfarlanes, Ashurst, CMS, Reed Smith, and PwC, utilize Harvey “daily” and that it is flying high. The blog post stated that since last December, Harvey’s yearly recurring revenue has increased threefold, and the number of employees has also tripled.