Soxton AI, an artificial intelligence-first law firm targeting early-stage founders and startups, launched today with $2.5 million in preseed funding.

The round was led by Moxxie Ventures, with participation from Strobe, Coalition, Caterina Fake and Flex.

Founded by Chief Executive Logan Brown, a Harvard graduate and former lawyer at Cooley LLP, Soxton seeks to democratize access to legal services for startup founders and sweep away the complexity of legal proceedings using AI automation.

“I’ve always lived at the intersection of law and technology… quite literally since I was in middle school,” Brown told SiliconANGLE in an interview.

Brown described how she was inspired by the movie “Legally Blonde” to explore the legal system, including seeking employment at her local district attorney’s office at the age of 12. She now combines that early legal experience with coding classes she took as a teenager, along with years spent steeped in the tech startup ecosystem.

To equip founders with better tools to tackle legal problems, Soxton blends automated workflows with a “lawyer-in-the-loop” model, offering fast turnaround and significantly lower cost compared with big law firms. Brown said she wanted to build a service where “a lawyer will get back to you within four hours,” with prices as low as $100.

Modern-day founding teams typically fall into two categories: first-time founders with limited legal experience, and repeat or later-stage teams that already use big law. First-timers can get guidance on incorporation, governance, equity plans and basic compliance to get off the ground rapidly. For veteran teams, Soxton handles the “low-hanging” repetitive work — automating drafting and review — which takes pressure off using big law, lowers costs and accelerates turnaround.

“It’s almost prohibitively expensive to use some of the bigger law firm options,” Brown said. “We want to allow more founders to get access to better services at an earlier stage.”

AI has become a sticking point for both lawyers and technology founders, especially when it comes to legal guidance. It’s not enough to drop into a generalized model such as OpenAI’s ChatGPT and ask for advice; numerous headlines have already shown how lawyers can land in trouble when an AI model hallucinates case law or misstates key details.

According to Brown, AI represents the first technological shift capable of meaningfully transforming a legal system that’s operated in its current form for more than 300 years. However, that transformation comes with pitfalls.

“Where many people make mistakes is that they’re using just ChatGPT… without having a human in the loop. And it’s not quite there yet,” Brown said. “These tools are designed to give you an answer, without concern for the nuances.”

When it comes to the legal system, it’s been operating for so long and with such a level of complexity, it’s difficult for everyday people to process. She stressed that Soxton’s vision is to provide “access to justice,” and do it in a way that won’t put users in legal jeopardy.

“There has always been an access to justice issue, and people’s ability to access the judiciary,” Brown added. “ChatGPT and these AI tools are great for overviews, but… it’s important to us that we have lawyers in the loop.”

The legal AI space is becoming crowded with legal AI technology firms looking to provide services, but most tools target lawyers. Examples include Spellbook, described as “Cursor for contracts,” Stockholm-based Legora AB, a provider of collaborative AI for lawyers, and Eve, a startup with an AI-driven software platform for law firms.

“We have a very long way to go before AI tooling can go directly to folks and provide accurate legal advice,” Brown explained.

With the new funding, Soxton will expand its operational backbone by hiring engineers, lawyers, data engineers and a dedicated operations specialist. For now, the company remains focused on helping founders at their earliest stages, but the AI-first firm aims to grow alongside them and become a larger part of their lifecycle as they scale.

Image: SiliconANGLE/Microsoft Designer