We’ve heard about Thomson Reuters, LexisNexis, and vLex providing genAI legal research tools, but Bloomberg Law wants in on this too. It has now enhanced its Answers feature to leverage LLM tech to tap the company’s growing corpus of legal data. (Plus, what does this mean for the wider market? See below.)
Its corpus includes: court opinions, practical guidance documents, and secondary sources such as books and manuals. This includes book collections published by the American Bar Association, James Publishing, American Bankruptcy Institute, ERISApedia, as well as books exclusively published by Bloomberg Law that are available only on the Bloomberg Law platform.
Clearly well aware of the ongoing accuracy debate, the company added that they provide ‘sophisticated guardrails [to] minimize the risk of incorrect answers and prevent the feature from answering questions outside the legal domain, while linking source documents for transparency and verification’.
They didn’t elaborate on which genAI models they are using for this.
Bloomberg Law Answers is currently available to users participating in the Bloomberg Law Innovation Studio, an ‘experimental environment’ available to select Bloomberg Law users. The Innovation Studio allows customers to interact with experimental product features and share feedback. I.e. this does not appear to be on general release yet, while given their corpus this is clearly focused on the US for now – although that may still be of interest to international law firms wherever they are based.
Todd Barton, vice president of product, Bloomberg Law, commented: ‘Bloomberg Law Answers represents a significant step forward in our commitment to integrating AI within legal research, in line with our long track record of leveraging innovative technologies to improve attorney workflows.’
—
Is this a big deal? Well, it’s perhaps better to say it’s a necessary move. If everyone is tapping their legal data with genAI then you’ve got to do it too. Simple as that. And it’s worth saying that Bloomberg Law is not new to using AI, they’ve been providing NLP / ML-based capabilities prior to this.
Perhaps the bigger question and one that takes into account other companies, from scale-ups such as Jus Mundi, to startups offering LLM-driven legal research such as Midpage, is how this will all shake out?
Some years ago it looked like the legal research market was all sewn up. Then vLex started to expand globally and bought up smaller peers, Wolters Kluwer likewise expanded its legal tech offering, Bloomberg Law also grew, and now we have small, agile genAI-first startups coming to market which tap free legal data and provide what they hope will be seen as more compelling UI/UX and workflows.
And we have even had rumours of Harvey seeking to buy vLex with a ton of VC money – although for now that one has gone a little quiet – although who knows if they will take a second run at it over the summer and match the price that vLex wants, which is also rumoured to be way higher than what Harvey and its backers were initially willing to pay? Harvey plus vLex’s vast body of already highly structured legal data would be a compelling competitor – if such a deal ever materialised. But, it’s best to see this as ‘informed market gossip’ for now and taken with a pinch of salt. Although, never say never….
In short, far from the legal research world having reached a ‘steady state’, it’s looking more fluid and dynamic now than it has for a long time, despite TR buying up Casetext in 2023 to take out a growing competitor with genAI skills. Of course, let’s not get carried away, the Big Two will continue to dominate for some time to come, but we cannot assume the market will always look like this, especially as genAI unsettles things and leaves open new market possibilities.