The AI Mirage in Law — Starting conversations

Jack Shepherd
8 min readMay 2, 2023

People working in software development are always on the lookout for ways they can optimise how they work. For example, linters are plugins that live inside a code editor that can tidy your code up as you write it. Coders working on front end code can use libraries such as Storybook, which helps to simplify collaboration by isolating user interface components from underlying business logic.

These are all things that bring improvement, but they are changes in how people work. As a result, they not only require upfront investment but also buy-in from the people affected by them. In most software development teams, sensible discussions take place around whether these kinds of changes are worthwhile. To the extent they are, upfront effort is prioritised in the same way as more exciting feature development.

Over the last decade, the number of technology tools available to lawyers has increased hugely. In addition, more and more of these tools affect high priority workflows actually conducted by lawyers. Technology has moved from the back-office to the hands of the lawyers themselves.

Despite this, it is realistic and fair to say that the majority of lawyers, law firms and legal teams do not operate in the same way as software developers when it comes to continuously improving how they work. Especially in larger organisations, sentiments such as “this is how we have always done it” prevail. In most legal jobs, people are so busy that client work will always take priority over non-client work.

While many are hailing the advent of generative AI (and other similar developments) as being a potential catalyst of change in legal, we shouldn’t forget that lawyers, as a whole, often lack the desire, time or space to even think about changing how they work. It seems a good time to reflect on the numerous factors that influence whether a particular tool will get adopted, and whether these factors will remain irrespective of any new developments in the technology space.

Quality of the software

When having conversations with lawyers about new technology, one thing that will nearly always come up will be the quality or intuitiveness of software they use. For years, there has been a huge gulf between the technology standards lawyers expect in their personal lives, and the standard they expect at work. For too long, people seem to have been willing to accept using software that looks at like it was made in the 90s, yet at home will complain when their self-driving car can’t drive itself down remote country lanes.

There are three things that have contributed to this issue in law firms: (1) infrastructure, (2) development issues, and (3) under-investment.

The first of these is likely to be addressed by firms moving to the cloud, rather than staying “on-premises”. Among other disadvantages, firms that do not make the move to cloud will have to manually update their systems each time an upgrade becomes available. Of course, every organisation has its own risk tolerance and policies in this area. But firms that do not move to the cloud are at a disadvantage in maintaining the latest and greatest.

The second factor, development issues, often manifests itself as lack of intuitiveness, difficulty to use or technology that is buggy or does not work. Software teams that do not follow best practices will continue to face issues here. In particular, a focus on product-thinking – which leads to truly understanding what people are trying to achieve in a particular area – rather than throwing coders at the problem, is absolutely key here.

One note of caution here – sometimes, software is said to be unintuitive because it does not enable people to do their jobs in exactly the same way they did before. This is normal, and indeed, if you do not get this kind of reaction to a new tool, you might wonder whether you are merely upgrading an existing bad processes, rather than fundamentally helping people do their jobs in a better way. In any case, it is the job of software teams and buyers of tools to ensure this change is managed properly.

The third factor relates to the first two – either lacking the time/resources to consider upgrading infrastructure or buying software based on price rather than quality. But most of all, under-investment in technology results from technology and process improvement not being a strategic priority of an organisation.

Of course, the recent surge in the interest of AI will cause law firms to refocus their strategic priorities to make sure they are not left behind. Some will be doing so on a “wait-and-see” basis, rather than because they have identified strategic priorities that can be honed by these new developments. And this is where we must start to talk about business models.

Business models

One of the reasons why software developers are so focused on continuous improvement is that their jobs depend on it. If they start to release buggy or ill-designed software, nobody will buy it, and existing customers will start to look elsewhere. Their business models depend on securing recurring revenue for their software, which depends on the value their customers are getting. Time and resource spent building the software might be reflected in licensing costs, but not directly; you are paying for the software, not the time spent building it.

It is certainly true that businesses will not instruct law firms that cannot deliver legal advice correctly, accurately or in a user-friendly way. Yet the business models of law firms and legal teams are not the same as in software. In many cases, you are paying for people’s time rather than the value of the product produced (the discussion around value of the product often manifests itself as a fee negotiation discussion).

Legal technology can still play a part here, of course. For example, if a firm does a particular kind of work but cannot do it within the budget clients have allocated to it, technology and process improvement can drive efficiencies to reach a mutually agreeable budget. For firms working on a fixed fee basis, this tools and efficiencies can be direct margin-makers.

However, it is apparent that in situations where law firms are delivering services to clients, and clients are paying the bills, the incentive to adopt new technologies is less clear. My personal view is that things like “enhancing the client experience” and “being efficient” are not sentiments that will generally overcome a lawyer’s natural resistance to change, because these things are not tangible business outcomes that leaders will drive them to deliver.

We shouldn’t blame lawyers for not being like software developers in this respect. I don’t agree with people who say lawyers deliberately take long over their work because it makes them more money – most lawyers are far too busy, stressed or working late to even start to think about that. I do not think lawyers should be criticised for continuing to work the way they do if there is no business or other incentive to change. It’s just that as long as they can provide good quality legal advice that clients are willing to pay for, lawyers’ jobs don’t depend on doing things differently. And therefore, the incentives to think about how they can improve are limited. It is difficult to see how technology by itself will change any of this.

It might take one or a handful of forward-thinking law firms to challenge its competitors through lower cost and better service. Or perhaps buyers of legal services instigate the change. (Most likely, both…).

Buyers of legal services

Many business have robust procurement processes for legal services, especially for firms wishing to become part of their panel. These processes often lay down requirements relating to internal efficiencies and financial reporting.

But in some occasions, these procurement processes have struggled to make an impact. If not run correctly, they turn into “check-box” exercises that both sides forget about once you are in the midst of an extremely busy project.

One of the issues here is that many buyers of legal services have themselves been lawyers in the past, and therefore accustomed to the status quo as far as working practices are concerned. For example, a transactional lawyer working in a law firm ten years ago would have been accustomed to the arduous process of managing closing checklists through a Word document, and spending hours in a PDF editor (or worse, a printer) sorting out various signature pages for signed documents. When this lawyer moves to corporate legal, they might accept that law firms have to do this process because they have been through it themselves. In this way, inefficient behaviours at law firms can become normalised.

Some in-house lawyers take a different view. Instead of allowing law firms to do it “the old way”, they challenge their firms to find a better way. In the example above, this might be through encouraging a law firm to use a transaction management tool.

But the reality is that clients do not always care about the internal tools their lawyers and law firms use to do their jobs. They might not see these tools as having any impact that is relevant to them. Especially at the higher end of the market – people might be swayed primarily by reputation and personal connections above everything else.

When law firms are doing press releases around the new tools they are using, they should frame these around the actual impact these tools are having. The impact is often “lower cost” for clients. Some firms might see this as a problem and hide from this fact. Those firms risk being accused of “innovation theatre” by focusing on other aspects which, while interesting, are probably not that important to clients. Others might have thought in advance about how efficiencies from technology can win them more work or help them become more profitable. These firms will likely end up with a more targeted and nuanced approach to technology deployment.

These are all business issues that go to the heart of the lawyer-client relationship. While discussions may be prompted by technology, they are not technological issues. This is why I cringe a little when I hear people say things like “AI is going to change the legal profession”. Without a fundamental change in relationship or business model, it won’t.

Category of software

The final factor that relates to all of the above is what a given piece of software actually does.

There are some technologies that law firms cannot operate without – such as email, time recording, conflicts management and document management. These technologies will not suffer the same adoption issues as those that improve workflows, but people have coped without for decades – such as transaction management and client extranets. For some firms, these are ultimately not things that a firm’s life depends upon – but are seen as “nice to haves”.

There is also an in-between type of tool – such as knowledge management and document automation, that fall somewhere in-between. Over time, tools will make their way from the “nice to have” categories to the “must have” category – but only if the extraneous factors identified above change this.

Again, the underlying technology underpinning a solution does not change this fact. It’s more important than ever to focus on what a piece of software actually does, and how impactful it is to clients and an organisation’s business model.

On the cusp of change?

While I do not agree with people who say “AI is going to change everything”, I still welcome them saying these things. That is because it starts a discussion. Once that discussion is started, it should be moved to the issues that matter, such as those identified above. The biggest mistake would be to keep talking about technology in isolation of the people, businesses and relationships it affects.



Jack Shepherd

Ex biglaw insolvency lawyer and innovation. Now legal practice lead at iManage. Interested in human side of legal tech and actually getting things used.