(Picture: Andrew Arruda is on the above left, alongside fellow co-founders Jimoh Ovbiagele, who is CTO, in the centre, and Pargles Dall’Oglio on the right).
Artificial Lawyer caught up with Andrew Arruda, the co-founder and CEO of ROSS Intelligence, the first AI-driven legal research system*, in order to ask one of the leading lights in the field of legal AI about the company’s meteoric rise.
The Origin Story
Artificial Lawyer first asked Andrew Arruda how ROSS got started. He explains that the company’s origin story stems from co-founder, Jimoh Ovbiagele. The need for a more efficient way of conducting legal work was made painfully apparent when Ovbiagele’s parents split up. The legal costs of divorce had been so high that the family didn’t have enough money to formalise their split. Clearly there had to be a better and more efficient way of providing legal services.
Ovbiagele also happened to be a computer programmer who started programming at the age of 10. On his LinkedIn profile he writes: ‘I want to use artificial intelligence to tackle humanity’s greatest problems. I cofounded ROSS Intelligence to create a better understanding of the law through artificial intelligence and human-centric design.’
Ovbiagele and Pargles Dall’Oglio, co-founders of what was to become ROSS Intelligence were working together at University of Toronto (U of T). Meanwhile Ovbiagele had met up with Arruda several times at various U of T events. Then, one day Ovbiagele reached out to Arruda who was at that time articling as a junior lawyer at a small law firm in Toronto, Azevedo & Nelson.
Was he willing to leave his desk in his law firm office to join a machine learning and natural language processing start-up instead? Yes, he was. And so ROSS Intelligence took a major step forward.
They moved into a basement in Toronto and the group initially focused on Canadian employment law as a subject matter to explore with their AI-driven research engine. The decision was later made to move to the US, specifically Silicon Valley, where the company remains based.
IBM’s Watson was also an important part of the company’s development. ROSS had access to Watson tech that is not always generally available, as U of T had partnered with Watson for the IBM Watson Cognitive Computing Competition.
Back in 2014 IBM had asked a number of teams from top universities to compete and show how best they could use Watson’s cognitive abilities. The ROSS project had been one of those of teams. IBM Watson remains linked to ROSS through a partnership arrangement.
However, although Watson has always been part of their ‘tech stack’, their own proprietary AI systems are what allowed ROSS to make the progress they have made so far. Arruda notes that using Watson allowed them to focus on building AI tech that did not exist, rather than focusing on what already did.
The nascent company then received investment from Y-Combinator in early 2015 and then further interest, support and investment from Dentons’ tech innovation platform, NextLaw Labs. The decision was also made to focus on US bankruptcy law as the main research area for the system.
ROSS and Arruda as CEO seemed to have hit the market at just the right time. The world in mid-2015 was ready for stories about artificial intelligence and ROSS provided it for them.
Nearly every major newspaper in the world seemed to have covered the story of ‘ROSS: The First Artificial Lawyer’. The company was off to a flying start.
Coming back to the present day in the Fall of 2016, ROSS has publicly announced six law firms using its system, with many more using it, but not going public. The most recent client win was Womble Carlyle, but other clients include Dentons and Latham & Watkins.
Although last year the main line for many commentators on legal tech was that AI would replace lawyers en masse, the reality is quite different from Arruda’s perspective.
He is very much of the ‘AI and lawyers’, not ‘AI vs lawyers’, school of thinking.
‘Human lawyers are at the centre of this. ROSS is not making judgment calls. AI is very powerful technology, but, it’s most effective when working in a team,’ Arruda explains.
‘We want lawyers to stop being data retrieval machines,’ he adds, something many lawyers will thank him for. Arruda notes that he also had to spend a lot of time doing such work when he was a practising lawyer in Toronto.
He then adds that the legal world is changing. It’s more competitive, there is more pressure on what is billable. ROSS and AI in general fits perfectly into this world as it provides very efficient levels of work at relatively low cost.
This in turn allows young lawyers to develop into the lawyers they want to be, i.e. advisers with great human skills, not people sat in front of a screen doing basic research all day.
‘ROSS will allow people to become better lawyers,’ Arruda says, killing off any suggestions that AI’s role will be to decimate the legal world. Far from it.
He mentions how in the 1990s there was a frenzy about how the internet would ‘melt our brains’. That said, he doesn’t see AI as a sideshow. It will be centre stage in the legal market of the future.
‘AI will be bigger than the internet. There will be a massive shift because of AI,’ he states.
He also says that AI will help with access to justice and explains how so much of the legal world is out of reach to normal people. ‘The users of legal services, they’re just a minority [because of the cost],’ he says, with perhaps the old memory of his friend’s family’s split-up resonating in his mind.
But, he is also a realist about why law firms will use AI. They will want to see a clear ROI.
Arruda states that it will make them more profitable and that a clear financial benefit will be – and already is – shown by using a system such as ROSS to handle legal research that otherwise the firm would have to cost into any work it did for a client. Given that many clients don’t want to pay for that work anymore, then finding a way to do it without deploying $180,000 associates is a sensible and economic step.
So, Where Next?
ROSS has proven itself to the market and in a very short period won several very large clients and backers. What now? Will the company remain only in North America and only focus on bankruptcy?
Arruda is enthusiastic about future growth. They will be releasing ROSS across a variety of other practice areas in the next few months.
ROSS is also being made use of within in-house legal teams and this hopefully will spread.
He’d also like to see the company venture into the UK at some point, but there are no firm plans yet to break into English law, though at least there would be some similarities given the use of Common Law.
Then Arruda says he’d like to see the company one day operate in Civil Law jurisdictions. This he adds is a future plan, but not one that is so far off.
‘Wherever there is grunt work to be done, there will be an AI solution,’ he states confidently.
And, given how much has changed in just the last 18 months then I am sure he is right.
[*Although, the company likes to point out that research is just the first step and that they are building an AI-platform for law in general.]