The Impact of AI in the Legal Industry

· Thomas Wood
The Impact of AI in the Legal Industry

Unlock Your Future in NLP!

Dive into the world of Natural Language Processing! Explore cutting-edge NLP roles that match your skills and passions.

Explore NLP Jobs

We examine the potential influence of machine learning and AI on the legal industry.

AI has transformed a number of industries but has not yet had a disruptive impact on the legal industry. It is unknown exactly how and when AI will become impactful in law, but it is important to prepare for this change.

Today, AI in law (legal AI) is now helping lawyers and entire legal departments do more, do it a lot better, and at a lower cost too.

AI and natural language processing in particular can be used to process contracts in the legal industry

AI in investigation, AI in crime, AI for legal – is this starting to sound like something right out of a sci-fi movie like Blade Runner or perhaps the futuristic dystopia portrayed in Minority Report? It’s ironic and quite befitting to say that movies do imitate real life, giving viewers a sneak peek into the future if not the present.

AI in law has only just begun to surface and come into its own, and lawyers across the globe are excited, anxious and even a bit nervous about what kind of impact this may have on the legal sector as a whole. It is probably fair to say that within the coming years, we may very well find ourselves right in the middle of a surreal revolution where law will be led by the adoption of AI, particularly, by in-house legal teams and lawyers.

Just like how email changed the way we communicate, AI in criminal investigation or AI in legal practices will become absolutely ubiquitous; an indispensable asset to almost every forward-thinking lawyer. And simply put, those who are reluctant to understand or shy to adopt this change will either get left behind or may no longer considered a “successful lawyer”. However, those that do adopt this change will ultimately find more time and energy to do namely two things which lawyers can’t seem to find enough time for: thinking and providing advice.

Before delving deep into the impact of AI in the legal field, it pays to first understand what artificial intelligence is at its core, in order to better appreciate how AI in the legal industry might revolutionise the way investigations are carried out or cases solved.

What is AI?

The terms AI or “artificial intelligence” tends to be a bit misleading at times – well, when it comes to its application in the legal sector, anyway. Let’s get this out of the way early on: AI in the legal sector certainly doesn’t mean a Terminator-like robot walking into court in a suit and tie, complete with a briefcase full of incriminating evidence – although admittedly, the idea does sound inherently cool!

Perhaps a better description of AI for the legal industry would be ‘cognitive computing’, one that lawyers are quickly catching on to. This essentially refers to computers having the ability to learn, communicate, reason and make decisions – without any human intervention whatsoever. These AI-powered cognitive tools can either be trained or programmed.

With the former, it involves learning to complete tasks that are typically done by people – where the focus is on looking for specific patterns in data, testing that data, and providing results. With the latter, you might think about it as a research assistant with the ability to sift through an entire deck of information and then reporting what it found.

So why should all this be important to the everyday lawyer or legal team? Because, according to an IBM report, 2.5 quintillion bytes of data is generated on a daily basis – that’s 2,500,000,000,000,000,000 bytes of data! In contrast, a human being’s ability to review and comprehend that volume of data is pretty much impossible (without any outside assistance).

Winner takes all?

In some industries, there has been the dynamic of a “winner-takes-all” scenario where it comes to data. Examples of “winner-takes-all” dynamics would be the tech giants dominating social media and online search. These are clear examples of network effects and economies of scale, where a business with large amounts of data can attract more customers.

The question of interest to a law firm considering adopting AI for legal purposes, is whether AI in the legal industry will lead to a “winner-takes-all” scenario, where large firms dominate the deployment of AI in the industry.

Disruptive entrants

There are a number of legal tech startups entering the market. Fast Data Science presented our Insolvency Bot at the JURIX 2023 conference in Maastricht, and we also attended the LEGALEX 2023 conference in London’s Excel Centre. We were able to see a number of fascinating emerging applications of AI, from law and philosophy to access to justice and to improving more routine tasks such as case management and bundle management.

The recent explosion in the popularity of AI across industries is due to improved computing power and machine learning algorithms, as well as increasing volumes of electronic data and falling costs of data storage. The legal industry, like healthcare and pharmaceuticals, has historically been conservative and highly regulated, and has not yet adopted AI to the same degree as in some other industries, such as transport, retail, technology, or hospitality, to name a few. However, that is about to change.

The legal industry can expect a move towards widespread adoption of AI and machine learning. In fact, according to experts, company spending on AI reached $47 billion in 2020 – that’s up by nearly 600% since 2016 when it was $8 million. Clearly, these staggering statistics point to one thing: AI for law is something that’s going to be very hard to ignore and live without by legal teams and departments everywhere.

The case for AI in Law

It makes sense to stop for a moment and think about this huge spending in AI: there are huge cost savings and productivity gains to be had, as well as the abovementioned access to justice. Namely, from freeing up people from routine tasks that computers can very easily handle – allowing people (lawyers) to focus more on tasks that add true value (better, more streamlined advice to clients) – which are typically things that computers are not good at.

This goal rationale fits all too well into the legal sector. But more importantly, legal departments must brace themselves for the change and quickly adapt to ‘AI for legal’. It’s not surprising to know that many leading universities including Stanford, MIT, Oxford and Harvard are adding MBA programmes introducing courses in AI.

The interesting thing to note here is that CEOs and CFOs have welcomed and embraced the change, and they will most likely expect C-Suite members to “follow suit”. This, of course, includes the general counsel and legal department. Therefore, in-house lawyers who are quick to embrace AI for law practices, will become an invaluable asset to the new generation of CFOs and CFOs.

So how does AI work or how might AI in law work, more specifically? Keep reading to discover the exciting possibilities AI in investigation and AI in the legal industry in general may have.

AI involves teaching a computer how to learn, perceive, reason, communicate, infer and make decisions, like real people do. The process initially involves something called machine learning (ML), where the machine – a computer – starts making decisions on its own with a minimal set of programming instructions to begin with. The interesting thing is, that instead of just writing rules manually on how the computer will interpret a specific set of data, ML algorithms allow the computer to determine all the rules on its own. In case you’re wondering, an algorithm is a set of instructions used to solve a particular problem or set of problems.

However, a much larger goal lies beyond machine learning: deep learning. Deep learning uses far more advanced algorithms to perform tasks which are a lot more abstract in nature, like recognising images, for example.

Ultimately, via machine learning, computers become better at what they’re supposed to do with experience. It’s a way of self-learning or cognitive computing, to be precise – kind of like the Skynet computer from the Terminator films, where the computer learns and adapts to such a degree that it starts controlling entire systems and infrastructures in the city, wreaking havoc in its wake. But we’re nowhere near that level of advancement and it’s probably just a concept befitting for a sci-fi thriller.

Fundamental to this self-learning process are three core mini-processes of how cognitive computing generally works:

  • Gather information;
  • Analyse to try and understand that information, and;
  • Make specific decisions based on the information.

As any lawyer would know from experience, this process in its very nature is iterative – i.e. we become better at something the more times we do it, particularly if we have someone overseeing our work and receive guidance and correction from time to time (characteristic of a young associate learning the ropes at a law firm, for example).

So, for the legal sector, it works pretty much the same way with AI.

Another aspect of cognitive computer-based learning has to do with the computer filtering certain things so that there’s no room for over-inclusive answers. For instance, a computer returning 5,000-10,000 results will certainly not be valuable to or usable by a lawyer. Therefore, the computer has to learn what’s relevant to the lawyer’s search criteria and then make suggestions which are usable in terms of narrowing down the results to a much more workable amount – say, a few responses, perhaps around 50-100. The lower, the better.

If all of this sounds too complicated or too good to be true, then consider the following real-world scenario where AI in criminal investigation was used to solve a murder case:

In late 2020, AI was used by the defence counsel in a high profile murder trial at the Central Criminal Court of England and Wales – commonly known as the Old Bailey.

James Watson, the accused, was due to stand trial in January 2021. The charge: murdering 6-year-old Rikki Neave in later 1994. The schoolboy was discovered without clothes, strangled in woodland, Peterborough, with his missing uniform dumped in a nearby bin.

Ruth Neave, the deceased’s mother, underwent a murder trial in 1996 and was convicted instead of getting a ‘child neglect’ sentence. Subsequently, she was given a 7-year prison term.

The crime had largely remained unsolved, and Watson continued denying the charges after the investigation was reopened in 2015. Through the course of the 8-week trial at the Old Bailey, Watson’s defence attorneys revealed that they had been using AI software to analyse over 10,000 documents in order to justify his innocence.

They claimed that the company which supplied them with the AI software dramatically helped to speed up investigative proceedings, including evidence examination as well as a search for specific patterns and connections which might have been overlooked during human inspection and investigative work.

While the findings or the outcome of the case cannot be revealed for legal reasons, Watson’s lead defence attorney, Sally Hobson, said that the AI-powered software provided important benefits and could revolutionise the way evidence is examined or investigations carried out.

Ms. Hobson revealed that she had the gargantuan task of sifting through several thousand pages of evidence and material, which she had to digest and make sense of in a matter of days; that the court’s existing digital case system had “massive limitations”, and the AI software learned what to look for, read and understand – and it yielded the required results within hours – what would have normally taken months to find. And that the accuracy of those results is also something that could save hundreds to thousands of pounds in taxpayer’s money.

Ms. Hobson added that the introduction of AI in law could speed up efforts to reduce the often overwhelming backlog of court trials which had piled up and caused undue delays due to the pandemic. She also said that while lawyers will continue to make key decisions on their own and that judges would still be ‘the judge’, everyone will be aided by AI technology to enhance and accelerate each case in truly amazing ways.

The technology used to determine Mr. Watson’s innocence was based on natural language processing (NLP), pattern recognition techniques and a mix of unsupervised and supervised machine learning – in order to analyse evidence and reach a swift decision.

Just to give you a quick example of how this would practically work: if you’re sifting through witness statements of people referencing that they saw a suspicious-looking man walking a dog – and you did a keyword search for “dog”, the regular search system might not provide the most accurate results, as it would not take into account similar search terms like “mongrel” or “mutt”.

With an AI-powered application or software, on the other hand, it will understand that the lawyer is searching for a man with a dog, and will pull all the applicable search results, and not just the ones containing “dog”. This may help find that golden nugget, so to speak.

The use of AI in the legal industry for case preparation is something that’s already being implemented in commercial cases. In fact, AI for law enforcement has already been deployed in a recent investigation against Rolls Royce by the Serious Fraud Office.

AI in criminal investigation is expected to become commonplace as the need to examine massive amounts of digital data for evidence on phones and other similar devices will become far greater. One major reason for integrating AI in law is its increased efficiency and speed, when it comes to examining data to find the desired results. Another one is having machines perform analysis which is actually a lot cheaper and potentially far more accurate than relying on people.

What we’ve discussed under this subheading alone forms the basis for the great leap forward in not only AI for legal departments but also AI for the legal industry as a whole – the ability for machines to learn tasks that could only have been done by lawyers, and the ability for lawyers to extract the exact information required by simply typing in a query or by asking the machine to perform a specific task.

AI for Law – Putting it all into context with practical uses

The technological innovations in AI can enable legal departments and entire firms to offer vastly improved representation to clients and optimise legal services. Here are just some of the ways in which AI for legal can really shake things up in the industry:

Access to Justice (A2J): a trickle-up phenomenon?

Article 6(3)(c) of the European Convention on Human Rights (EHCR) guarantees the right to legal aid where a defendant does not have the means to pay for it, or where it is in the interests of justice. In practice, in the UK and elsewhere, access to legal advice is often the preserve of the wealthy. In the UK, legal aid was cut back so much that criminal barristers went on strike across the nation in June 2022, as many were finding legal aid was insufficient to cover their travel costs.

For this reason, the legal industry is one where AI is making headway in lower-ticket cases such as immigration and insolvency law. At JURIX 2023 we saw a fascinating presentation on AI for insolvency and bankruptcy in Finland. Domains such as family law, immigration law, and bankruptcy and insolvency, are areas of law where litigants and defendants often find it hard to access legal advice around the world. For that reason, it is quite possible that many legal AI innovations will come out of the Access to Justice (A2J) movement, and possibly “trickle up” to bigger ticket areas of law, such as international contract law.

Insolvency and Bankruptcy advice

Together with an international team at Royal Holloway, Gunnercooke OPL and University of Surrey, we have developed a question and answer bot which can answer questions about English insolvency law, using Retrieval Augmented Generation (RAG). The tool uses relevant text of key UK statute law such as the Insolvency Act 1986, important case law from the National Archives, and information on procedures from HMRC’s website, the system triages incoming queries and sends a smart and informative prompt to a generative model. You can read more in our accompanying blog post.

Fast Data Science - London

Try the insolvency bot

Corporate insolvency RAG demo, presented at JURIX in Maastricht. Not legal advice.

Click here to download the slideshow presented at the JURIX 2023 conference.

Our paper was published in the proceedings of the 2023 JURIX conference in Maastricht, the Netherlands. DOIDOI10.3233/FAIA23097910.3233/FAIA230979

Ribary, M., Krause, P., Orban, M., Vaccari, E., Wood, T.A., Prompt Engineering and Provision of Context in Domain Specific Use of GPT, Frontiers in Artificial Intelligence and Applications 379: Legal Knowledge and Information Systems, 2023.

Contract reviews

Part of a law firm’s work involves reviewing contracts for their clients to take apart any terms which could potentially have negative consequences for them. AI and ML technology can effortlessly review documents and contracts to find any risks, suggesting changes to help clients get better terms. With machine learning at the core, the software will only get faster, smarter and more effective after each use.

Conveyancing and land transactions

We were contacted once by a potential client in Australia, who wanted to deploy a large scale AI solution to handle land titles. There is a particular kind of land title document in Australian law, which often needs to be quickly parsed to find the names of certain guarantors.

Similarly, we have worked on a project in a consulting role for a national Land Registry, developing a natural language processing identifying sections of the land title free text which correspond to polygons in a map file.

Predicting case outcomes

One proof-of-concept which we were asked to investigate at Fast Data Science was the possibility of using machine learning to identify criminal cases that are likely to lead to a successful prosecution, for a major public body. In short, criminal cases come with a number of witnesses, witness statements, ages of witnesses, and other information. The likelihood of a successful prosecution varies by the type of court and type of crime. LexisNexis has a legal analytics spinoff called LexMachina which claims to predict case outcomes using historical court judgements.

Better profitability

AI analytics is something which can have a positive impact on almost every aspect of a legal firm, including accounting, advertising and client procurement. What the human eye fails to catch, machines can succeed in rapidly detecting patterns across massive amounts of data in order to establish certain connections and trends, and even inconsistencies.

Spotting such patterns in advance can enable legal firms to delegate assignments in a way that ensures exemplary legal services. For example, machines can help collect all the necessary information, evaluate the data after collection, and at the same time also discover ways to help legal firms run more profitably and efficiently.

Improved productivity

The majority of legal work can be very monotonous and cumbersome – AI can improve productivity to a great degree by automating tedious day-to-day tasks which don’t really require any expert input. This way, AI can enhance the efficiency and accuracy of legal procedures by automating a number of menial tasks which may otherwise and unnecessarily waste time as well as resources. Lawyers in the UK are all familiar with online resources such as LexisNexis, which use natural language processing to assist in information retrieval.

Automated text and image redaction, NLP for anonymisation

Redaction is the removal of text and/or images from an original legal document by blacking out specific words, phrases or entire paragraphs of text or pixelating images. However, redacting not only requires lots of attention to detail but also multiple pairs of eyes to review and scrutinise the final document. Errors can easily slip by owing to the sheer volume handled by lawyers each day – and despite their best efforts to protect and conceal sensitive information, it is very hard to eliminate errors altogether.

AI-based redaction software can produce fast and easy redaction of audio, video and PDF files by using pattern-based automated redaction. And, by automating the redaction process this way, law firms can effectively cut costs by a significant margin – as manually redacting records is not only a costly practice but also very time-consuming.

Eliminating research errors

Part of a lawyer’s daily essential practice is conducting legal research, but lawyers are human and there is a risk that they make an error or overlook something. AI-powered software can help eliminate these errors completely by producing the desired outputs far more rapidly and efficiently.

Machine learning algorithms, for example, are capable of finding documents and data very quickly which may be relevant to a specific case. They also have the ability to highlight changing legislature over the course of a set timeframe and how they may vary from jurisdiction to jurisdiction. This saves law firms a lot of time as the collected information is virtually free from errors and does not require to be scrutinised again by a lawyer.

Client acquisition tracking

When combined with marketing data, machine learning can make it a lot easier to make sense of client acquisition and retention – including the specific ad campaigns which gained the most popularity and how certain clients chose one law firm over the other. This data can help provide a clearer understanding of which law firm performs the best, as well as the lawyers with the best success ratio.

In Closing

Artificial intelligence is enhancing the legal profession in a number of ways and just some of the advantages of integrating it include improved profitability, optimised time management and access to more relevant information.

Fast Data Science’s artificial intelligence and machine learning consulting services will help you come up with a tailored legal AI solution to help make your internal processes far more efficient and cost-effective. Call now for an initial consultation: +44 20 3488 5740, or send us a message.

Your NLP Career Awaits!

Ready to take the next step in your NLP journey? Connect with top employers seeking talent in natural language processing. Discover your dream job!

Find Your Dream Job

Generative AI
Generative ai

Generative AI

Generative AI Introduction Generative AI, a subset of AI, is fundamentally transforming industries and shaping the future. Leveraging advanced algorithms, generative AI can create content, designs, and solutions that were previously unimaginable.

Big data
Big data

Big data

Big Data The emergence of big data has revolutionized industries, transforming traditional business models and decision-making processes. In this comprehensive exploration, we delve into what big data is, its significant impacts on business strategy, and how companies can leverage vast datasets to drive innovation and competitive advantage.

AI in finance
Ai in finance

AI in finance

AI in Finance The integration of artificial intelligence (AI) into the finance sector has revolutionized how institutions operate, from automating operations to enhancing customer engagement and improving risk management.

What we can do for you

Transform Unstructured Data into Actionable Insights

Contact us