Hello, Legal World. It's Me, GenAI!

How generative AI is poised to shake up the legal ecosystem.

The legal industry is firmly in the crosshairs of generative AI. A recent Goldman Sachs study found AI could automate up to 44% of current legal tasks and vertical AI startups like Harvey are starting to reveal how. As a former “cladiator” at Ironclad and future Wachtell attorney, I’m seeing a shift beyond traditional research, contract management and e-discovery solutions, to more “core” lawyer activities like practice management and intellectual property management. While historically lagging in tech adoption, the legal world is now at an inflection point driven by the increasing prominence of Chief Legal Officers (CLOs), growing compliance burdens, and talent shifts away from Big Law among other trends. In my view, vertical AI will be the catalyst for legal's long-awaited digital transformation. In this roadmap, I’ll unpack the market drivers for why legal tech is booming, share specific predictions for the future, and a handful of startups to keep your eyes on.

Why now?

1. More work ⬆️ 💼 , less fees ⬇️💰: It’s not just that AI is here. It’s that law is ready. Lawyers (both in-house and at firms) are under pressure to do more with less. Roughly ~70% of legal departments are seeing higher matter volumes, yet 66% report flat/declining budgets. Not only are case volumes increasing, but also the modern lawyer’s responsibilities extend beyond traditional “legal” work to include strategy and operations. As an example, in 2023 former General Counsel Jason Kwon was promoted to Chief Strategy Officer of Open AI. Lawyers routinely influence the direction of how products can and should be sold, made, provisioned, regulated, monitored. As legal reasoning is increasingly viewed as integral to businesses across sectors, tech adoption will rise.

2. Unlocking domain-specific data 🔓💻: High quality data is the foundation of all successful LLMs. Legal work is data-rich, but it lives in analog. Bringing legal data online will unlock tremendous application potential. Domain-specific LLMs in fields such as legal and finance have demonstrated that building niche foundational models outperforms generalist models in their respective fields. Initiatives like Kelvin to build a specialized corpus of legal documents will enable enterprise AI offerings to flourish.

3. Talent shifts 👫✈️: Top legal talent is leaving traditional Big Law. Instead, they are pushing innovation at startups that unbundle legal services. This influx of domain expertise, and the emergence of repeat legal tech founders, is poised to fundamentally reshape legal workflows and services. Lawyers are ready to experiment.

Four predictions for the future

1️⃣ The rise of the CLO/GC stack. ⚖️

Behind every growth-stage CFO is a robust software stack that helps them track and manage key financial information across their business. What about CLOs? Until now, most software designed for in-house legal teams has focused on simplifying contracts and document management. But companies like Streamline AI have shown that legal works on much more than just contracts. As the office of the CLO becomes ever more interdisciplinary, software tools that help not just with contracts, but also with legal research, reporting, compliance, spend management, and intellectual property are becoming “need-to-haves”. In addition, the current cost-cutting sentiment is driving the convergence of legal tech with fintech and regtech. Startups like Norm AI and Lexverify that focus on easing compliance burdens and Oddr which simplifies the legal billing lifecycle are solving “hair on fire” problems for modern legal teams. As enterprises embrace legal technology, client pressures will subsequently drive adoption of legal tech within law firms. Integrations with existing tools, data privacy, and context-specificity will be critical.

2️⃣ From generalized copilots to specialized agents. 🤖

As GenAI-for-law approaches become more sophisticated, I’m also seeing a shift from general-purpose copilots to more specialized products that can assume an agentic role. That is a shift from AI that helps lawyers to AI that actually performs legal work. The unbundling of traditional firm services with solutions like Litera (Kira) reveals that not everything lawyers do is necessarily legal work. When considering tasks typically billed at law-firm rates, LLMs excel at summarization, fact finding, contract review, and document categorization. Startups like Harvey and Macro are removing the manufacturing aspects of law and freeing up lawyer time to focus on judicial and strategic acumen that is core to the field. While labeled as "legal assistants," platforms like Paxton and Legalyze.ai are executing traditionally billable work. As others like Bench IQ work to understand and systematize seemingly unknowable processes like judicial decision making, automating legal reasoning seems more possible. AI-native vertical startups are reasoning machines that are ushering in a new era for the practice of law.

3️⃣ Trailblazing across sub-verticals within law. 🔎

Specialty areas of law are commonly overlooked - both by practicing attorneys and by tech. As legal data goes digital, a wave of startups enabling broader access to more niche legal services will emerge. Today, most of the action is in intellectual property. Companies like Solve are using AI to help attorneys write patents and realize the full value of inventions. Outtake.ai is automating legal actions against IP harming content. But I’m also excited about the widening scope of vertical AI to other legal functions such as tax, divorce, mass torts, immigration, and public interest law. For example, Darrow is building a justice intelligence platform that connects public interest attorneys with high-impact cases. Just like the CLO stack, bringing data online with high context-awareness, interoperability, and accuracy will be key.

4️⃣ Legal empowerment for all. 🦾

Finally, I’d be remiss to overlook consumers and small businesses - the ones without Big Law spend budgets. For them, interacting with the legal system involves substantial cost and knowledge barriers. In many cases, this is necessitated by the economics of law firms, and somewhat justified by the immense amount of knowledge and expertise required to be an effective legal professional. However, startups are identifying processes that can be effectively handled without the expertise and time of a lawyer and productizing them to make them dramatically more accessible. Think: immigration, wills, divorce, and more. For example, Casehopper and Lawfully are using AI to simplify the immigration application process. DoNotPay offers subscription-based online legal services. The common thread is the identification of specific legal services that (a) are commonly needed by those outside the legal system, and (b) require repeatable, clearly defined processes that rely on unstructured but consistent information.

Across the legal tech landscape, there is work that requires trained lawyers. And clearly, work that doesn’t. Vertical AI will make this distinction clearer. Looking ahead, I’m excited about innovations that not only streamline legal processes but also reshape the very essence of legal practice itself! If you’re working in this space, I’d love to chat - you can find me at lakshmi@siliconbench.com.

Introducing the next generation of companies bringing GenAI to law:

Vertical GenAI x Law Market Map