The last few years have seen increased speculation about the impact of artificial intelligence on the legal sector and, in particular, on the timing of that impact. Tony Williams looks at the challenges and opportunities presented by the new technology
A great deal of technology is currently available which will help to make the provision of legal services more efficient, whether that be project management tools, document drafting programmes, e-discovery processes, or document management, to name a few. This technology is becoming both more user-friendly and cheaper, and is now therefore accessible to a wider range of law firms.
Law firms of all sizes need to understand what is happening around AI and to adopt change in a way that is appropriate for them and their clients. Wait and see may become wait and die!
But artificial intelligence (AI) is a significant advance on existing systems because it is capable of learning from the processes it performs in order to provide a better and more complete service the more times it does a particular task. This is potentially a major differentiator from where we are now, and it will create a more dynamic and rapidly changing environment when it happens. AI is intrinsically linked to ‘big data’, since AI relies on the quick, comprehensive and accurate analysis of billions of pieces of data, which no human could ever have the time or capacity to collate and evaluate.
In this article, I look at how AI is and could be used in the legal and other sectors, and the challenges and opportunities it creates for lawyers and legal businesses.
In the US, IBM Watson has been primarily focused on the medical profession. Many leading doctors have been working with it to help diagnose particular conditions and to then develop the appropriate treatment regime. Watson is able to analyse the results of research papers, drug trials, patient data and other sources to provide the health professional with probabilities of certain diagnoses and suggestions of optimum treatments. All of this is based on a detailed analysis of hard data. The more Watson deals with certain conditions, the better it gets.
The potential benefits of this approach are significant: a shorter time to diagnosis, less intrusive and expensive investigations, optimised treatment plans, an accurate prognosis and greater patient confidence. At a time when health budgets worldwide are under pressure, the potential cost savings could also be substantial.
Interestingly, many health professionals are embracing this development. They see it as a valuable tool to help them provide a more accurate diagnosis and treatment plan. However, at this stage, it does not usurp their role as professionals, and it is unlikely, at least in the short-term, that patients will wish to dispense with the services of a senior doctor.
A potentially more disruptive use of AI could be in the fund management industry. An AI programme could potentially analyse any event, from a change of interest rates or a change of government to a natural disaster, map this against, say, 1,000 companies, and analyse its impact on their supply chain, operating costs, sales, cash flow, effective tax rate and so on; this would give a fund manager an important advantage over managers who do not have access to such data. However, it may not be necessary for even the fund manager to analyse the data. Instead, based on its understanding of the implications of the event and the likely level of impact on the relevant companies, the AI programme could be enabled automatically to buy or sell shares in the relevant companies. Given that speed of execution is already crucial in rapidly moving markets, if this technology is developed, it will quickly move from an interesting to a business-critical tool. This may present challenges for fund managers, who may struggle to justify their roles; many so-called active funds rarely achieve consistent returns in excess of passive index-tracker funds, despite the higher fees that active fund managers traditionally charge.
AI could affect a wide range of professional services. It could in fact have an even larger effect on the professional sector than automation had on manufacturing.
But what does this mean for the law? A significant amount of legal work is about collecting and collating vast amounts of data, whether evidence, statute law, case law or market practice, and then applying an expert analysis to it and recommending particular courses of action.
In many respects, the lawyer’s role is similar to that of the doctor mentioned above. In the US, there are already systems available that analyse historical data on a firm’s record before a particular judge, to analyse the likelihood of its success or failure in future hearings before that judge. A data-driven analysis of the likelihood of success or failure based on an extensive review of past decisions is likely to provide a party with a far more accurate indication of the certainty of the likely outcome. Clearly, if both the claimant and the defendant have access to such systems, they will understand just how large the ‘zone of uncertainty’ is in relation to any dispute. This may well encourage more realistic attempts to achieve an early negotiated settlement, whether by mediation or otherwise.
In areas of contract negotiations, a system that reviews thousands of signed contracts can help to identify the usual outcomes – the negotiated compromises that are made – and assist in the development of a negotiation strategy. Clearly, certain matters will have different dynamics, and the client may have different priorities, but a clear understanding of how similar deals have developed will be an important tool for lawyers. Currently, prior executed documents may not be in the public domain, although an increasing number are. However, a review of the law firm’s own database (suitably anonymised to protect client confidentiality), or of the client’s own records (which will demonstrate what it accepted in the past) would be a valuable exercise in developing both the first draft and an acceptable final version.
The effective development of AI could also have a serious disruptive impact on professional services, including the law. In our research report, Civilisation 2030: The Near Future for Law Firms, published in November 2014, we looked at a number of long-term trends and their likely impact on the law (a copy of the report is available to law firms upon request to Jomati). One of these was AI. We chose 2030 because we felt that by that point AI would be having a major impact on the legal profession.
Currently, it is very difficult to predict the pace of development in the legal sector. We are traditionally a rather conservative profession, reluctant to embrace major change. The profession has changed and continues to do so, but this tends to be an evolutionary, rather than revolutionary, process. Furthermore, despite the increased pressures on law firms since the downturn, law is still a very profitable business. Accordingly, the usual conditions for major change are not currently particularly evident in the legal sector. But as technology develops and specific products gain more profile and acceptance, the pace of change will almost certainly increase. Increased client demand for lawyer efficiency, price transparency and the predictability of outcomes will force a degree of change.
A range of law firms are experimenting with or developing a variety of expert systems (although, currently, they may fall short of being defined as AI). This experimentation and development will continue and accelerate. Indeed, many firms claim that with the help of technology and low-cost centres, they are currently achieving a higher profit margin for routine work than on supposedly high-value bespoke work, despite providing a lower price and often better quality product for the client.
So how can law firms of all sizes respond to the challenges and opportunities that technology and potentially AI represent? There are a number of steps that can and should be taken.
Many law firms’ existing systems often have a range of useful functions that are rarely, if ever, used. In some firms, multiple products have been purchased with similar functionality. A proper audit of your current systems, what functions are used and by whom, can result in major savings on licence fees, and also provide further functions at no investment cost (except perhaps some more advanced training).
There is now a plethora of technology solutions providers in the market offering a vast array of products. In some cases, they seem to be offering ‘solutions’ to problems that we did not know existed!
In such a dynamic market, some products will work well and others will not. Some will be suitable for your client and business mix, others will not. Rather than sign up for firm-wide products, experiment with suppliers of new products in a specific area or office. Review progress against the supplier’s promises, and if the product isn’t working for you, reject it.
But always learn from the experience. This may be challenging in a law firm where many lawyers only want to act if the outcome is sufficiently clear and any failure is unacceptable. However, this is a period of experimentation. Some firms should even build a research and development function and budget to ensure that they remain in a strong position. This does not mean that a firm has to be a ‘first-mover’. For many, not being on the ‘bleeding edge’ of technology may be a distinct advantage, but this is not an excuse to be a late adopter, either.
Regrettably, many law firm partners are not prepared to consider change. Don’t try to convert them all, but work with those facing the greatest need to change – for instance, following panel reviews or in response to client demand for cheaper or more efficient service. If you succeed with them, then others will want to come on board.
None of us has a monopoly on the ‘right’ answers, so a collaborative and learning-based approach is likely to be more efficient and cost-effective. This may take lawyers out of their comfort zone. However, many in-house legal teams have a limited IT budget and would welcome an opportunity to partner with their law firms on these issues.
The changes produced by technology and AI can be daunting. I suspect that a large number of lawyers hope that they will have retired before its full impact is felt. Although the pace of change is uncertain, the direction of travel is pretty clear. Law firms of all sizes need to understand what is happening and to adopt change in a way that is appropriate for them and their clients. Wait and see may become wait and die!