Interest in Legaltech and AI has reached new heights, but unlike previous hype cycles, lawyers and investors are seeing real impact

At conferences, CLE sessions, demo days, and LP meetings, discussions of AI in the legal space are ubiquitous. Founded on the promise of LLMs and fueled by sky high valuations for companies like Casetext and Harvey, buzz around the application of AI to the legal space has reached a fever pitch.

Legaltech has been hyped in the past, only to come crashing down without delivering on its promise. In 2020 at the height of the last bubble, Atrium raised over $75M to disrupt law firms only to fold a year later. But while these historic failures offer a number of important lessons, I’m convinced that today’s investment will leave an indelible mark on the legal space — particularly in the world of Intellectual Property.

To better contextualize the current legaltext boom, I spoke to VC investors who have tracked the space for the last decade. Many of them contributed to, and even lost out from previous hype cycles. They unequivocally believe the legal space, and particularly IP, is on the precipice of lasting change. “The advent of Microsoft Word made work significantly faster and let lawyers focus more on other more strategic and impactful tasks. I think we’ll see the same thing with AI. Workflows will change radically, giving IP departments time back to focus on generating profit.”

AI will unlock value for IP teams, but beware of noise

One consistent takeaway I heard from every investor I spoke with is that these levels of hype inevitably create noise. This could be companies trying to “AI-wash” their products by adding AI to marketing material without changing their underlying technology. It could also be highly technical teams building AI tools without the necessary insights into how legal and IP teams operate. Worst case, it could be companies that sacrifice proper protection for speed in a space where data security is critical.

The biggest concern IP teams should address with AI is confidentiality. Says Sam, “Asking the right questions when reviewing new AI tools is critical. Software built by teams specializing in IP with in-house IP attorneys will go to great lengths to make sure using their AI can’t be considered a public disclosure. However, it is a great question to ask.” In addition to confidentiality, teams should ask providers about data use, security, and ownership as well.

Sam also advises teams to make sure the AI they’re implementing maximizes profit potential — i.e. that your AI helps you get the most out of your IP you can, on top of saving you time and energy: “AI is finally allowing IP departments and tech transfer offices to transform from cost centers into solid profit centers, a huge shift. To do this, AI-native platforms must successfully take the extremely opaque and arduous process of monetizing IP and make it 10x easier and faster. Because this is happening, I think every IP producing company will have a focus on monetizing their IP.”

Jessica from Emergence encourages IP teams to have proactive conversations with AI vendors about how else they can benefit from AI.  “Think beyond what vendors are offering you today to what you’d like to see them build for you. Communicate those needs so the tech becomes true leverage for you, not just a commoditized nice-to-have.” Legal departments have heard false promises about legaltech, and even though AI is now advanced enough to help IP teams with tasks like monetization, results across platforms will vary. Don’t let all the noise around AI for intellectual property make you miss out on the opportunity to start working with the right tool for transforming your IP strategy and operations.

So what makes today different for them? Unsurprisingly, many VCs cited the maturity of AI, both in terms of accessibility, sophistication, and privacy. But many also pointed to much deeper changes in the legal space. Corporate legal teams are increasingly trying to claw back work from outside counsel. But as budgets tighten, the only way to keep pace with their workload is to adopt tools to drive efficiencies across already lean teams. VCs also keyed in on the transition of the IP team from a cost center to a value driver.

Let’s dive into the specifics of how the potential for legal AI to transform IP has evolved into what it is now.

Why this cycle of AI buzz isn’t just hype

AI has come a long way — It’s orders of magnitude better than it was five years ago. But how do investors define better? First, the underlying models themselves are inherently better at completing legal tasks. Beyond that, deployment models have changed considerably. Today, SaaS startups can deploy incredibly powerful, cloud-based LLMs into legal departments with relative ease. Finally, the ability of startups to incorporate unique datasets and deep know-how into LLMs without years of training is a game-changer.

Model Sophistication: In 2017, a paper published by Google called “Attention Is All You Need” opened the door to LLMs by creating a mathematical model for word meaning. Before that, it was difficult for AI to understand word order, which meant it couldn’t properly structure sentences either. Law is all about language, so legal AI before LLMs was challenging. Now, LLMs let computers interpret and generate text with human proficiency.

Deployment Models: Implementing AI tools in legal departments used to require bespoke data science pipelines, robust data cleaning, and unique front-end development for each deployment. This meant that most AI companies targeting legal in the last decade were really services companies, not tech platforms. While the concept of an AI Agency was not without potential, it lacked the scalability and economics that make SaaS such a desirable model.

For Jessica Cohen at Emergence Capital, the ability to deploy these models as powerful SaaS tools can reshape IP portfolio management: “For the longest time, IP has been managed as a static asset. Once AI has been applied and portfolios are finally seen as evolving assets that can rise in value, that unlocks so much more. Namely, commercialization, which will have an impact that goes way beyond the legal world. Making sense of existing tech across different industries and companies will unlock the distribution of knowledge and skills like never before.”

Incentives are aligned. Perhaps the most well-documented barrier to AI adoption during previous bubbles is the law firm billable hour model. Pitches for significant efficiency gains often fell on deaf ears at firms for whom productivity gains meant fewer billable hours. Early success of companies like Harvey that target large law firms suggest these customers are finally evolving to accommodate demand for flat fees or other alternative fee arrangements.

Even more important: corporate legal and IP teams have spent the last five years trying to reduce their dependence on outside counsel. For these customers, promises of efficiency gains and cost savings are incredibly poignant, especially now that legal budgets have begun to contract and efficiency has become paramount.

Sam Garcia, an investor at AMPLO VC, equates productivity gains from AI to the impact of modern productivity tools like Microsoft Office.