Upcoming WebinarsMore Webinars
Oge Marques explaining recent developments in AI for Radiology
Author of the forthcoming book, AI for Radiology
AI Business is part of the Informa Tech Division of Informa PLC
This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 3099067.
In their 2017 report on the sector, PwC claim that clients have long been frustrated with the rising cost and speed of legal service delivery and many have responded by taking more work in-house, by bringing individual lawyers in through ‘lawyer on demand’ providers. Many believe that the time is ripe and that the “2020s will be the decade of disruption” as law firms increase investment in technology to automate tasks and improve decision making.
“It is clear that AI, and more specifically ML, is set to disrupt the legal profession. Anyone who thinks the profession is immune is in denial," argues Ben Allgrove. Allgrove is a partner at Baker Mckenzie, one of the largest global law firms. Baker McKenzie are adopting AI today in 11 offices over 3 continents , as part of a world-wide roll out plan to transform the speed of document analysis. Allgrove has a broad range of AI expertise, thanks to an academic background in the application of legal and ethical systems to artificial intelligence systems. He believes that the goal of all lawyers, whether in-house or in private practice, is being a trusted advisor to a client. Ultimately, embracing technology in the right way could help firms achieve that.
Disruption will not remove the need for lawyers but will instead usher in a new breed, Allgrove argues.“There will be a rise in the lawyer who can leverage technology and bring legal domain expertise to clients as part of wider, more holistic solutions than is common today”. With this, the industry will move into the twenty-first century - but he admits it will require organisational, as well as cultural change. "Law firms as well as in-house teams are going to need to become much more resilient organisations. They need to be built around lawyers bringing technology-enhanced skills to the table, with less of the current focus on delivering legal knowledge.”
AI will be used to deliver efficiency gains, what he calls “our version of Henry Ford’s ‘faster horse’”. However, the most interesting opportunities for him pertain to how the technology can be leveraged to deliver quality gains – “new insights, better risk judgements, better prediction, better outcomes”.
“The winners in the next 5 years (and beyond) will be those lawyers who embrace AI and truly start delivering ML enhanced judgement to their clients. The tipping point in the profession will be when the fastest horses start racing cars."
In order to prepare for this wave of disruption, and ensure that the technology is applied in a meaningful way, Allgrove believes it is important to address the main misconceptions surrounding AI today. For example, many are fearful that the technology is “prime time ready” to replace jobs overnight.
Although there are examples today of solutions applied very well to specific tasks, like eDiscovery and due diligence, overall the main market problems are far more difficult to identify and even more challenging to solve. “Two main technical/commercial challenges exist”, Allgrove argues. “a lack of data and the investment required to train the systems to provide client-tailored solutions.”
He admits it is going to take time and investment to develop the applications of AI in the legal profession, particularly at the top end of the market, where the phrase "good enough" still translates to "perfect" in most clients' minds.
It's vital that to challenge the idea that we are inevitably headed for the age of the robot lawyer. He believes that although AI is set to disrupt the profession, the human element will always be needed.
“Yes, there will be some, indeed many, things that humans do today that they won't do in the future. But the practice of law is more than just about the task, or even knowledge of the law. It is a suite of skills, which include judgment and, most importantly, empathy.” The latter, Allgrove argues, is at the core of what it means to be a trusted advisor. "In this way, human in the loop is vital when it comes to legal sector”.
AI won't solve every problem we have – just look at the current pitfalls of document automation.
“Our experience of contract automation is that it works brilliantly for simple, short, standardized documents. However, once the baseline document becomes more complex and the upfront human input requirement is increased, automating the process actually becomes less efficient than a traditional precedent doc with drafting notes.”
He argues that “more tech is not always going to be the answer”. The key, Allgrove reveals is about identifying the main problem and “then deploying AI where it enhances the outcome and not just deploying tech for its own sake.”
Allgrove believes that for business leaders looking to implement AI, success will depend on getting priorities right. “At present, my main priority would be getting the infrastructure right. It is essential you get your data organised; get it into the cloud; understand what you have, and most importantly what its limitations are.”
With the right infrastructure in place, business leaders will be in a good position for the wider deployment of the technology, Allgrove says. He believes that that second focus should be on people and talent. “Disruption is here and it will force organisational change. Therefore, the main questions facing leaders are (1) what workforce and skills are you going to need (2) How are you going to train that workforce?” The firms and in-house teams who can effectively crack these pressing questions will have a distinct advantage moving forward.
One major talking point surrounding AI today is the over-reliance on opaque and potentially biased mathematical data models. The argument goes that if bias inside training data for algorithms that make ever-important decisions goes unrecognised and unchecked, it could have serious negative consequences, especially for poorer communities and minorities. Added to this is the fact that technology companies are secretive about how their algorithms work and rarely disclose information about how decisions have been made.
Allgrove believes this is a huge question the sector needs to face up to as it brings together two of the big topics in discussion around regulating AI – bias and transparency. “I would say that the objective – whether in legal AI systems or anywhere else - should be to make sure we understand the biases in the data on which AI systems are learning and to affirmatively and transparently decide whether we care.”
The problem is that to produce a completely unbiased system requires a completely unbiased dataset. “That is an ideal that does not exist in the real world. So better to accept that bias exists and seek to identify and understand it. After this, you can then correct for it, either in the development process or when using its outputs (ideally both). Indeed, there are some fact patterns where you may positively want to insert a bias into the system. The debate about affirmative action to tackle gender or racial biases is a classic example of that.”
Ben Allgrove is taking to the stage in New York and is presenting on the impact of AI and the legal sector. “I am looking to explore what it means to be a trusted advisor in an AI/ML enabled world. Are clients ready to put that sort of trust in a system? Does it matter if they do? And how trust is going to be a key determinant of where AI most impacts the legal profession. I am fascinated by what AI is going to be able to do in 5+ plus years' time rather than in what it can do now. Show me the future!”
Author of the forthcoming book, AI for Radiology