As a lawyer in a mid-sized specialist firm, I recognise a lot of this. We use Harvey but most of the tasks it really excels at for us are things that I suspect we will eventually be able to get an Azure-based agent to do for us: run this specific playbook (that we own) and give me a review checklist, search this repository of documents to pull out who the directors are of every company listed, review this document and hand off the extracted data to our document production platform, etc.
For firms like ours, I expect it will be smaller job-specific AI agents integrated within existing (non-AI) workflows where the initial productivity gains will be.
Great article, as a lawyer I now understand why Harvey's valuations are going crazy (was scratching my head many a time). Also, many Magic Circle firms foresaw a while back that they could just develop their own AI legal tech - many have established their own Knowledge and Innovation departments since a few years back.
This analysis hits the mark, venture incentives are shaping legal AI more than lawyer needs. The real winners will be the products that solve genuinely hard problems and empower lawyers, not just sell fear or rely on hype.
Nice analysis! I think one of the other problems with LegalTech companies is that they're too broad. They're trying to becoming the next ChatGPT for lawyers but instead of that startups have to go deep and obsess over solving one problem for a customer. For example, in IPR patent research is a very tedious task and a tool can be built for this or a copilot specifically for transactional law. This will enable two things: 1) Higher Domain-Expertise in the tool 2) Fewer Hallucinations. I write on LegalTech as well. Let's connect
Great analysis. I think this even applies to AI companies outside the legal tech space too. I’m keeping my eye on these companies.
Good stuff, Jordan. We featured you guys in our just launched personal injury newsletter: https://newsletter.rankings.io/p/airlines-face-new-scrutiny-over-toxic
As a lawyer in a mid-sized specialist firm, I recognise a lot of this. We use Harvey but most of the tasks it really excels at for us are things that I suspect we will eventually be able to get an Azure-based agent to do for us: run this specific playbook (that we own) and give me a review checklist, search this repository of documents to pull out who the directors are of every company listed, review this document and hand off the extracted data to our document production platform, etc.
For firms like ours, I expect it will be smaller job-specific AI agents integrated within existing (non-AI) workflows where the initial productivity gains will be.
Great article, as a lawyer I now understand why Harvey's valuations are going crazy (was scratching my head many a time). Also, many Magic Circle firms foresaw a while back that they could just develop their own AI legal tech - many have established their own Knowledge and Innovation departments since a few years back.
This analysis hits the mark, venture incentives are shaping legal AI more than lawyer needs. The real winners will be the products that solve genuinely hard problems and empower lawyers, not just sell fear or rely on hype.
Nice analysis! I think one of the other problems with LegalTech companies is that they're too broad. They're trying to becoming the next ChatGPT for lawyers but instead of that startups have to go deep and obsess over solving one problem for a customer. For example, in IPR patent research is a very tedious task and a tool can be built for this or a copilot specifically for transactional law. This will enable two things: 1) Higher Domain-Expertise in the tool 2) Fewer Hallucinations. I write on LegalTech as well. Let's connect