Applications of Artificial Intelligence in Law Firms and ChambersReading Time: 5 minutes
The concept of artificial intelligence has tantalised humanity for decades. While AI was once merely the subject of golden-age sci-fi novels and films of wildly varying quality, today AI is permeating through ever-growing areas of life with advanced feats of machine learning closer to being a reality than ever.
While we’re still a far shout from Asimov’s robot masters ruling the universe, the applications of artificial intelligence in the modern day have the potential to revolutionise work in law firms and chambers across the UK.
The role of artificial intelligence in law today
There are numerous ways in which parts of the legal sector are already implementing artificial intelligence. The speed and accuracy with which a machine learning algorithm can perform specialised tasks has the potential to outstrip that of human employees. LawGeex, a legal tech company, has been developing AI to perform complex, time-consuming legal tasks such as analysing NDAs.
Whereas twenty different lawyers took an average of 92 minutes to complete five NDAs, the AI completed it in just 26 seconds. On top of that, it achieved an accuracy rating of 94%, matching that of the best-performing lawyer in the sample, who on average achieved just 85% accuracy.
Of course, there are plenty of other companies developing AI for the legal sector. The German company Leverton has developed a cloud-based tool which speed-reads documents in twenty languages. eBrevia, founded by former junior associates, has an algorithm that gathers relevant information from different documents. Then there’s Kira Systems, whose due diligence software apparently makes first-time users up to 40% faster, rising to 90% as they gain experience with the algorithm.
The clear trend is that artificial intelligence in law is being implemented to cut out the huge turnaround times associated with many aspects of the legal profession. Rather than outright replacing people, AI is being used to improve user performance by accelerating output while improving accuracy.
The future of AI in the legal sector
With numerous businesses investing in research and development, the prevalence of AI in our future seems like a sure thing. So how might AI continue to benefit the legal sector in the future?
The increased accuracy and efficiency offered by AI could potentially alleviate the difficulties of an overburdened legal system, and could also prevent discrimination against sections of society disproportionately accused of crimes.
However, facial recognition algorithms (a highly publicised aspect of machine learning) have been found to support racial biases, as they struggle to identify people of colour compared to their ability to recognise white individuals. So if AI is to eliminate discrimination in the courts, it is imperative that the teams developing the software and the firms using it be as diverse as possible.
In 2017, Facebook shut down two of its experimental chat-bots, Bob and Alice. This was because they had been communicating with each other in their own language. While this was incomprehensible to their human handlers, it enabled the two bots to work more efficiently together. This ability to analyse, develop and comprehend complex linguistics are precisely what would make this sort of algorithm effective at sifting through complex legalese and even translating it in layman’s terms.
But what’s really fascinating about this is the effect AI could have on legal jargon as a whole. If Bob and Alice can streamline their communication by inventing their own language, machine learning could redefine legal language as we know it.
Will AI replace lawyers?
This is usually the first thing people think of when you ask how AI might affect society. But will the robots really take our jobs? Neil Sahota, co-author of the book Own the A.I. Revolution: Unlock Your Artificial Intelligence Strategy to Disrupt Your Competition, think it’s a possibility. Sahota pointed out in an interview that: ‘It’s predicted AI will eliminate most paralegal and legal research positions within the next decade.’
But others don’t think human lawyers will disappear anytime soon. Research by the McKinsey Global Institute found that only 23% of tasks in the legal sector could be fully automated. They concluded that while more tasks in the future could be performed by AI or non-lawyer personnel, lawyers will still be required for more cognitive tasks at the higher end of the legal profession. The idea that AI will go on to enhance the performance of human workers rather than replacing them isn’t uncommon.
However, it follows that by increasing efficiency, you reduce the need for staff as we know them (and their function) to be. Lawyers perform numerous essential functions for their clients, and it’s hard to imagine them disappearing anytime soon. But as Sahota suggested, the pressure that AI will take off of paralegals and the like in terms of workload could mean that there will be a changing in function and focus for staff across the next 5, 10 and 20 years.
How Weekly10 leverages AI to improve employee engagement for firms and chambers
Machine learning algorithms of varying complexity are already being used by organisations across the world for all sorts of purposes. We at Weekly10 are no exception. We have already incorporated artificial intelligence into our proprietary employee engagement software. Our custom-built and trained machine learning algorithms are prevalent in three key areas of the platform:
- Sentiment analysis
- 10Pulse – our employee engagement metric
- Smart question edit suggestion
Sentiment analysis for clearer employee engagement data
Put simply, sentiment analysis allows us to gauge how employees feel about a given topic based on the language used in their check-ins and feedback responses. We take the quantitative data provided by staff in their weekly employee check-in, and our algorithm, trained on more than 10 million data points gets to work. Sentiment analysis sets itself apart from the more common ‘keyword analysis’ by taking context in to account when assigning meaning.
10Pulse: a science-based approach to employee engagement measurement
Developed in partnership with academic partners, 10Pulse is our very own employee engagement metric calculated from all data supplied by employees based on the 5 key pillars of great employee engagement:
- Discretionary effort
- Job satisfaction
- Feeling valued
With 10Pulse, clients can track trends in engagement not only across the whole firm but at location, department and team levels, to get a clear view of how engaged their staff are (and in what areas they may be falling short).
Get smarter with your check-ins with suggested question edits
Question suggestions are one of our latest features in which a machine-learning algorithm uses information from previous reviews and check-ins, the algorithm can suggest improvements to existing questions and even entirely new ones. It understands the different types of questions that might be asked, what their purposes are, and how they might be framed more effectively.