Head in the cloud: getting started with AI
Artificial intelligence has burst into the legal services sector and it’s now here to stay, says Stephen Brown as he shares some practical pointers to get you going
According to PwC’s 22nd Global CEO Survey, 85 per cent of business leaders believe AI will significantly change the way they do business in the next five years, with nearly two-thirds seeing it as a bigger force than the internet. AI can work in your firm by reducing risk, improving efficiency, or allowing your legal teams to be more productive.
RISK, EFFICIENCY, PRODUCTIVITY
Let’s start with cost management, which is one of the main causes for complaint. AI can be combined with predictive analytics to analyse all the matters within your financial management systems, and report whether or not this matter will complete within the original costs budget. Simple solutions available today can deliver a similar function, but an AI solution can process more data and perform more ‘what if ’ scenarios. For example, an AI solution could consider the complexity of a case by reviewing documents in the data management system (DMS) for key phrases, and historical performance of this team and fee earners to manage costs effectively, to analyse the burn rate of this case and cases which match the matter profile. Next, improving efficiency, which AI can do by removing the monotonous tasks. Consider the flow of information (data) in your firm – where do you have teams of people reviewing information (data) and performing action-based logic? Areas where this exists are candidates for an AI solution. For instance, one firm used to employ a team of people to manage enquiry mailboxes. The task involved reading the email and assigning the enquiry to the appropriate team. An AI solution was implemented to complete the job, which allowed the team to focus on more value-add tasks. Another example is using AI to extract key information from documents stored in a DMS and automatically profile these correctly. This delivers an increase in ROI for your DMS, as search yields improve results and the ability to improve knowledge across the firm. A third application is in respect of productivity. Reviewing documents and completing an action based on the content of that document is a staple activity of law firms. There is a large number of document-based AI solutions that allow legal teams to be more productive, such as reviewing NDAs, litigation discovery, contract and lease review.
SECURE AND ROBUST
Historically, the challenge for law firms was identifying a provider or having the financial resource to embrace emerging technology. Financial barriers are now being removed, and there has been a rise in the number of suppliers delivering solutions. The challenge now is focusing on a specific issue that can make a real difference. In this regard, the Cloud plays an important role in the AI world, which it is important to understand before embarking on your own AI journey and asking some important questions. Not all AI systems are held in the cloud. It is acceptable and not uncommon, for a firm to have local AI services, but this relies entirely on the in-house computer hardware resource, capability and security. The preferred AI deployment model is cloud-based, not least as the space and operational systems that support AI are generally bigger than most in-house IT systems. Also, by definition, the data AI systems need to do what they do, must be substantial. The majority of AI providers run their services via the cloud, and are probably making use of data centres provided by Microsoft, Google or Amazon. There has been an explosion in the number of AI providers that are running their services in this way, but some AI providers have built their own cloud. As a user of AI, you shouldn’t need to care whether the cloud behind the chosen system is holding large or small amounts of data. What you need to know is that it is secure, robust and will meet all the key requirements you specify.
BE CURIOUS
The numbers of AI providers and suppliers has rocketed in the past year or so. To find the one which is right for your firm, you will need to create a shortlist of providers, talking to IT, business and innovation experts for their recommendations. The selection process starts with compiling a list of questions: 1 What do we need and want AI to do? Which area/s of the law do you want AI to support? This should come from identifying your firm’s challenges and opportunities. For example if there’s the potential to save time and generate fees from rapidly checking and updating commercial property leases, then this should definitely feature in your ‘AI brief ’. 2 Where will my data be stored, and who has access to my data? This is important to ensure compliance with data protection rules and SRA requirements, so it is vital to establish the jurisdiction in the world where the data will be held. 3 Which cloud provider will be used? This would be either one of the big data centres mentioned above or an ‘own cloud’ system. 4 How long have they been an AI provider? An established player will have been around for at least three years. Any less than this and there could be risks attached, especially if this is a new provider to market. However, as an emerging technology, it is to be expected that there should be a steady stream of new entrants.
BE PATIENT
From a standing start, it takes between six and 12 months to appoint an AI provider and for the chosen provider to get up and running with an AI system ready to work. l Allow three months for the initial engagement stage. l Once selected, allow time for your chosen AI provider to trial the system. This early stage is entirely dependent on what you want AI to do and the tasks you wish it to carry out. It is also critically important to the performance and impact of AI on your firm, because AI needs to be taught and to learn. l So, your engagement model will need to factor in time for your AI provider to curate your data, audit it, and ask the right questions to make your AI teaching steps align with your objectives and desired outcomes. The quality of the data provided in the early stage is absolutely key to deliver the results you want. l As with all IT systems, there will be a bedding-in period. AI will need to gain your trust and confidence so in the first stage of implementation, it may be sage to run manual checks alongside the AI processing to ensure the findings and actions are the same, and accurate. Over time, this will scale back so the level of manual checks will diminish.
WHAT TO EXPECT
Currently the top areas in which AI is most widely seen is, perhaps unsurprisingly, those with the greatest volume of documents and data, such as contracts, property leases and standard document review. These specifics also represent those for which AI delivers the greatest savings, in people, time and costs. But with its capability to consume data and analyse what it finds, AI also delivers vital insights into law firms’ clients. The contribution it makes to sentiment analysis for marketers, reputation managers, and custodians of your firm’s brand and proposition, will continue to grow, because it provides instant information about perceptions and feelings for the firm and its work. And that impacts and helps shape the strategies surrounding the firm’s CRM and overall client satisfaction. But the AI explosion won’t stop there. AI is already featuring on tenders and pitches for legal work. What was once an optional, nice-to-have add-on for some businesses seeking to engage a forward-thinking law firm will become a critical and must-have function in the next two years. As to the landscape itself, it will continue to evolve. There will be growing numbers of niche AI providers specialising in certain aspects of legal service, whether it be commercial property leases or NDAs for example. AI truly represents a valuable tool that enables repetitive, regular legal work to be handled faster and more efficiently, today, but even more so tomorrow. For legal professionals, it could be described as the path to liberation, which frees them from the timeconsuming, but essential, tasks, so they can invest more time in the high value, higher earning work.
