Ex Parte, a ‘Moneyball for Law’ startup, has bagged $7.5m in Series A funding. It says that it can ‘forecast the outcome of cases with approximately 85% accuracy’ – which this site has to say seems doubtful in all but the simplest of cases, but is happy to be proved wrong.
The US company, which got going in 2017 and has four staff listed on LinkedIn, also claims to be ‘the world’s first company to leverage AI to predict the outcome of litigation’. But, we also have Premonition, which makes…(er…made?), very similar claims in the past and officially started back in 2014.
Then we have Gavelytics and Solomonic, which both provide litigation analytics that help lawyers, but are at pains to stress they cannot predict the results completely, nor are they hoping to. They aim to give a steer based on past data, that’s all.
And we have the work of people such as Ludwig Bull and team, who have developed a number of companies with litigation prediction at their centre. They are more bullish, if you’ll excuse the pun, but where they have worked in relation to litigation finance the focus has been on smaller and simpler matters, and even then the litigation finance fund has still used plenty of its own human input.
In short, software can help, for sure, as it can crunch lots of prior data and see patterns – (if that data is relevant and in a form that allows such crunching to take place, which it often isn’t) – but basing litigation outcomes on pattern spotting is a tricky business. And that’s why firms use this kind of tech as a statistical overview of possibilities, rather than base their litigation strategy wholly upon it. E.g. you can get a steer on which judges allow which kind of motions in relation to certain types of cases. But, total outcome prediction? Very few people claim to be able to do that (…with good reason).
So, what does Ex Parte have to say? This is how they explain things: ‘Ex Parte is the world’s first company to leverage artificial intelligence and machine learning to predict the outcome of litigation and recommend actions that its customers can take to optimize their odds of winning.
‘The company’s patented Prediction Engine forecasts the outcome of cases with approximately 85% accuracy, and its patented Recommendation Engine generates data-driven recommendations such as whether to litigate or settle, which claims to assert, where to file or defend a lawsuit, and which attorneys and law firms will provide customers with their best chance of success given their case-specific factors.’
‘The Company will use the capital to expand its engineering team and invest in sales and marketing initiatives to further scale new customer growth. Ex Parte’s customers and prospects include corporations, hedge funds, law firms, insurance companies, litigation finance firms, and universities,’ they added.
The CEO is Jonathan Klein, who was the Chief Legal Officer for Microstrategy, a BI company – so, you can see the connection.
He said: ‘There are many technical and conceptual challenges associated with building a robust model of a lawsuit. Legal data is by its nature disparate, unstructured, and semantic. We have solved these problems by combining a highly specialised understanding of the legal field with advanced expertise in artificial intelligence, machine learning, and natural language processing.’
‘Our mission is to be the leading worldwide provider of data-driven decision-making solutions in the field of law and provide our customers with a winning advantage. Think of us like Moneyball, but for a market more than 20x the size of Major League Baseball.’
The funding round was led by R8 Capital and Ironbound Partners.
Overall, more litigation data analysis is welcome, but perhaps they need to manage expectations? The challenge is that ‘Moneyball’, which is a great book and movie, doesn’t easily fit with the law. There are several reasons why this is so, and a whole article is needed, but in brief:
- Despite what people would like to think, the law isn’t computer code – see article.
- While gathering data sounds likely to help, the question is: what data do you have? Do you actually have enough case law related to damages claims against auto companies for accidents caused by faulty headlights in the State of Illinois, to provide any real insight?
- And is that data useful in itself? Just tonnes of case law is just words and words. It’s all got to be taken apart and filtered. If you use NLP then this is a big task in itself. Not to mention issues such as weighting, and how you classify and develop outcome patterns.
- Disputes are super-complex and have many very human moving parts that are hard to estimate using case law.
- Moreover, a lot of commercial litigation never reaches a courtroom, so your data set is limited again.
And the list goes on.
In conclusion, litigation analytics is a very real area of legal tech growth. It’s also one that has come to understand its limitations. So, it’s good to see Ex Parte bag some cash to help it to grow, but if investors are expecting literally Moneyball for Law, then they may be surprised.
Great comments here on the limitations of predicting case outcomes. In particular that all those settlements are not actually data. What makes a case settle – with what outcome – is the larger territory. Perhaps the tech should be aiming at that?