Robert Ambrogi Blog
Last December, LegalOn, one of Japan’s largest providers of AI contract review technology to corporations and law firms, announced its expansion into the U.S. market, offering limited early access to its product to select users and organizations.
Today, LegalOn is making its platform commercially available to all customers in the U.S., and as it does, it is releasing a new feature, AI Revise, which it says is the “first AI contract editing tool enhanced by expert legal knowledge.”
“Our AI Review assists legal teams by spotting risks in third-party contracts and redlines and suggesting solutions to make contracts stronger before signature,” said Daniel Lewis, LegalOn’s U.S. CEO. “Now, we’ve taken a leap forward – AI Revise enables users to turn our suggestions into revisions.”
Read more about LegalOn in the LawNext Legal Technology Directory.
As LegalOn comes out of its early-access period, it is introducing several new features in addition to AI Revise. They include:
- Coverage of more contracts, including master service agreements, purchase agreements, non-disclosure agreements, and a variety of general provisions that are likely to appear in any type of contract, with its AI trained to spot risks specific to each contract type. Additional contract types will be added over time, with coverage of SaaS agreements coming in May.
- An in-browser editor providing the ability to edit contracts and collaborate with colleagues within a document.
- The ability to upload and search prior clauses, contracts and templates to find on-point language as you draft and edit.
- A comparison functionality to compare two contracts article-by-article to see differences between them.
Last week, Lewis, Mike Contillo, head of product, and Corey Longhurst, head of growth, gave me a demonstration of AI Revise. A key feature of the product is that it can be used right out of the box, with no setup or training required. LegalOn has pretrained the AI to fix hundreds of specific contract risks for every type of contract it covers.
“At the very highest level, what we’re trying to focus on as a company is helping organizations do pre-signature contract review, helping them negotiate stronger contracts faster by helping them find and fix contract risks,” Lewis said.
AI Revise is the company’s next step in pursuing that mission, he said.
“Working with generative AI, we are uniquely positioned to deliver value for customers because we’ve been doing this for the last six years. We’ve already built a very robust AI infrastructure to review contracts and we’ve built masses and masses of legal content to help people fix those issues once we spot them.”
One-Click Revisions
What AI Revise does is allow someone who is reviewing a contract to adopt LegalOn’s expert-developed content and suggestions with just one click.
On the left side of the screen, the user sees the contract being reviewed, and on the right side, the user sees the risk alerts that LegalOn’s other product, AI Review, has generated from its review of the contract.
“These are created by LegalOn’s AI infrastructure, as well as warehouses of expert legal content that we’ve been writing and refining through our in house team of expert lawyers over the past several years,” Contillo told me. “This all comes off of the shelf — no implementation, no playbook required.”
When the user selects an alert on the right side of the screen, the user is taken to that point in the contract in the left screen. The user can also expand the alert to show more information about the nature of the problem and the recommendation for how to solve it, including sample language.
What AI Revise adds to this review is the ability to click and have the AI made direct edits and redlines to the contract to address the issues raised by the alert, with deletions shown in gold-colored strike outs and additions shown with blue underlining.
Notably, AI Revise is not simply substituting boilerplate or templated text. It appears to be revising the contract in a manner that both resolves the alert and that is congruent with the context, structure, and formatting of the contract (including numbering and capitalization).
Once AI Revise makes its edits, the user has three choices. The user can accept the edits and move on to the next alert. The user can reject them by clicking undo. Or the user can tweak the revision by directly editing the contract in the browser view.
“The revision is being generated with the context of the actual original contract article in mind,” Contillo said. “And not only is it considering that context from a substantive level, but even from a structural level as well. What we’re excited about is that we’ve been able to successfully have it generate output in a manner such that the legal terms are harmonized.”
The user editing a document can tag others within an organization to collaborate. When that second user goes in, the user will be able to see prior edits that were made and by whom they were made.
Expert Guidance
Contillo believes that a key differentiator of LegalOn’s review technology is that it is informed by expert legal guidance developed over a number of years.
“The really important thing is it’s the first to work right off the shelf,” he said. “You don’t need a playbook. You don’t need implementation. The user doesn’t have to do preparation in order for it to work. There’s no work required from the customer. And the volume of potential tailored output isn’t constrained by what someone has pre-programmed.”
Of course, there will be companies that do have pre-existing playbooks and that will want to include those in the training of the AI and the recommendations it makes.
Lewis said that a problem with playbook-based editing is that it just swaps one clause for another. “The level of granularity that we’ve built is really unique, and that’s because we’ve invested the time to build out the content and the guidance alongside the software.”
That said, LegalOn will, within the next couple of months, launch functionality that will allow organizations to add their own preferred clauses and have them appear alongside LegalOn’s guidance and alerts.
Lewis declined to provide explicit pricing for the platform. He said that the company offers various pricing options contingent on types of contracts, numbers of users, and volume of use. All of the options include AI Revise.
“We feel confident that we’re coming in with a day-one usable option that can generate great ROI based on the price points that we’ll be offering,” Lewis said. “It should be immediately value accretive to organizations.”
While LegalOn is new to the U.S. market, it is a market leader in Japan, where it previously operated under the name LegalForce. The Japan company has more than 3,000 customers, 400 employees and $131 million in fundraising, including $101 million raised in 2022.
Lewis, who LegalOn tapped to lead its U.S. expansion, was previously cofounder and CEO of the legal research and analytics platform Ravel Law, which LexisNexis acquired in 2017. After the acquisition, he oversaw several LexisNexis businesses, most recently as vice president and general manager of practical guidance and analytical.