Legal work is incredibly labor- and time-intensive, as cases should be pieced together from vast amounts of evidence, leading some firms to conduct AI testing to streamline certain steps. after According to a 2023 American Bar Association survey, 35% of law firms now use AI tools of their practice.
OpenAI-backed Harvey is one among the large winners within the emerging AI legaltech space, alongside startups like This here and Klarity. But there’s room for yet one more, say Jerry Zhou and Kyle Lam, co-founders of an AI platform for private injury law called Spiritwhich emerged from obscurity on Tuesday with a $25 million investment led by Sapphire Ventures.
Supio uses generative AI to automate bulk data collection and aggregation for legal teams. In addition to summarizing information, the platform may also organize and discover files — and snippets inside files — that may be useful in outlining, drafting and presenting a case, Zhou said.
“After attending quite a few conferences and meeting tons of of lawyers across the U.S., Lam and I made a decision to deal with personal injury and mass tort law,” said Zhou, who can also be CEO of Supio. “These are practice areas that require compiling hundreds of documents from various sources and analyzing the information they contain and extracting information from them.”
Zhou and Lam are childhood friends whose skilled paths often crossed. The pair worked at Microsoft, specifically within the Office 365 organization, and in addition together on the tax software company Avalara.
The idea for Supio got here about after Zhou and Lam left Avalara to construct an organization that, in Zhou's words, “could help understand complex data and discover critical relationships inside given data.”
“We selected the legal industry because we knew it was not only document-heavy but additionally required technological innovation,” Zhou said. “(These are) areas of practice that require compiling hundreds of documents from various sources and analyzing the information they contain and extracting information from them.”
Personal injury and mass tort lawsuits, or civil suits on behalf of victims harmed by negligence, equivalent to the sale of defective products, are typically filed in the shape of documents equivalent to medical records, police reports, insurance claims, financial reports, consumer complaints, etc. Supio, Zhou said, creates demand letters — letters outlining the legal disputes to be resolved — in addition to supporting documentation, while allowing users to browse the evidence through a chatbot-like interface.
This could be very harking back to EvenUp, a startup that uses AI to create legal documents for assessing claims. Companies like Lawyaw and Atrium are also using AI to create initial claims.
However, Zhou claims that Supio is more complex in its technical approach.
“Law is amazingly complex and nuanced, and most productivity tool developers lack a real understanding of the legal documents that lawyers ultimately have to provide, which hinders the event of accurate (AI) models,” said Zhou. “Supio runs tons of of models with different features concurrently to try to know and classify documents. We then compare this to the expected work results and step by step improve those results.”
AI like Supio's is a robust tool in theory – nevertheless it's also a difficult subject. Given the sensitive nature of most litigation, lawyers and law firms could also be hesitant to grant (or be prohibited from granting) a tool like Supio access to case documents.
Late last yr, the California Bar Association released Guidelines that instruct legal professionals to not enter client data into AI tools that “lack adequate or sufficient security.” (Zhou says Supio stores client data in its home country and uses security protocols that comply with data protection regulations equivalent to HIPAA and GDPR.)
Another problem with AI Legaltech is the Perjured tendenciesLast yr, a bunch of lawyers from the law firm Levidow, Levidow & Oberman, PC tapped ChatGPT, OpenAI's AI-powered chatbot, prepared a private injury lawsuit against an airline. The result was disastrous: ChatGPT fabricated citations, misidentified judges, and referred to airlines that didn't exist. The federal judge overseeing the case, imposed a $5,000 high quality for the lawyers and their employer.
Courts are frantically preparing for a surge in inaccurate filings on account of using AI-based legal research tools.
In November 2023, the U.S. Court of Appeals for the Fifth Circuit suggested a rule (since scrapped), which requires that any skilled filing of legal documents created using artificial intelligence must certify that a human has reviewed and approved the documents for accuracy. Earlier that yr, a district judge in Texas issued an order Ban using generative AI to draft court documents without human fact-checking.
The risks of AI are so great that in a current Opinion poll Of over 300 legal advisors and senior legal advisors in large corporations, 25% said they imagine their outside advisors shouldn’t use AI. A separate Opinion poll by Thomson Reuters found that one in five law firms have issued warnings in regards to the use of AI.
Zhou makes the remarkable claim that Supio's AI achieves “higher accuracy than humans” and is “without hallucinations,” meaning it never lies.
“Supio provides flexible software with AI that may organize unstructured data and deliver reliable results because we all know that timing and accuracy are critical,” he said.
It is just not clear what is supposed by “human-level”; Zhou didn’t provide any test or benchmark results. But I might note that simply because AI can achieve feats equivalent to passing the bar exam, it doesn’t necessarily have the talents that lawyers acquire through experience and training. (The National Conference of Bar Examiners argues (a lot.) As for the “no hallucinations” a part of Zhou’s claim, that too is just not supported by any data—at the least not any that Zhou volunteered.
Nevertheless, some corporations imagine that Supio is promising.
Zhou said Supio currently works with around 30 personal injury and mass tort law firms and expects that number to grow to 100 by the tip of the yr. The startup's annual revenue is now over $1 million, with most of its money coming from subscription fees that Supio charges based on caseloads.
It may very well be about maintaining with the neighbors.
In a Opinion poll In a survey of legal professionals published by LexisNexis, just about all (90%) said they expect their investment in generative AI to extend over the subsequent five years; the identical survey found that 43% of firms now have a dedicated budget for generative AI. Gartner predicts that the fascination with generative AI will increase the worth of the legaltech market to $50 billion by 2027, almost double the market value in 2022 ($25.6 million).
Against this dramatic backdrop, Seattle-based Supio continues to operate. The company employs 27 people and expects to double its headcount in the subsequent 12 months.
Now that Supio has raised a complete of $33 million, Zhou says he intends to expand his client base within the near future – and eventually “expand into other legal areas.”
Bonfire Ventures and Foothill Ventures also participated in Supio's latest tranche (Series A). Zhou says it was oversubscribed but declined to reveal a valuation.
