GenAI Has Split the Legal Information Atom – Now What?

Generative AI has ‘split the legal information atom’ and unleashed an incredibly productive force. We now have the power in our hands to truly reshape the legal world. But will it happen?

Just a Convenience or Systemic Change?

Are we going to wield the capability of LLMs to really make a change to the way the legal sector works – in both meanings of the word ‘work’?

Or, are we going to be satisfied with saving an hour or two, here or there, every week on tasks that don’t change ‘the standard model’ of our time-based, high-leverage legal economy?

While busy announcing ‘game-changing’ technology (and it will be if we really follow through……) do we risk instead just going on to perpetuate a system that is soaked in profitable inefficiency and opacity?

And it’s worth adding that this seemingly immutable system receives an all too common indifference on the buy side of the legal sector equation. This is primarily because the buy and sell sides of the commercial legal world are in reality far closer to forming a self-perpetuating closed circuit, rather than a truly dynamic market for legal services…(for now).

Will we re-design whole practice streams from the ground up, using new models for legal productivity, or will we keep generative AI as ‘a little helper’ that won’t have a sufficient economic impact to alter how either law firms or inhouse legal teams view the business of law or how value is created and measured? (As has happened with so much legal technology before now and even with new working methods such as ALSPs.)

If it’s to use genAI mainly as ‘a little helper’, often on edge cases, e.g. drafting a client email, or summarising the occasional piece of new legislation, and it stays that way, then it would be a squandered opportunity, albeit with the most impressive technology the legal world has ever got its hands on.

While if it’s the former, i.e. to make a conceptual leap as to how the legal sector could operate, then what Artificial Lawyer has been working towards since this site started may finally be more than a possibility, and instead start to become something probable, perhaps even inevitable…..if initial conditions are met (see below).

But, let’s be open about this: it will sometimes be a bumpy road, for sure, and not all aspects of genAI are a positive, especially in the creative world, but….for legal, a world immersed in factual text, complex written contexts, and tons and tons of data, all underpinned by immense volumes of human activity, much of which is not high value, nor especially enjoyed by those who carry it out, yet manifests itself as business friction and an economic brake on society’s full potential….then this really is a huge opportunity for change. We just need to grasp it and act.

Where We Are Now

At present, genAI is steadily seeing uptake mostly across the larger law firms and major inhouse legal teams. And that’s a great start – it’s also highly predictable that it would begin there.

So far, from what Artificial Lawyer (AL) has seen and heard after talking to plenty of people, genAI is often being used in an ad hoc way, rather than with a team-based, process-planned, ‘total work product’ approach.

And by a total work product approach, that is to say a group of lawyers asking: what are we making here? How do we make it? How can we make it differently and better using all the new methods that are available to us? How do we then create re-usable models and designed pathways, as well as capture intelligence and knowledge for future use, so we can repeat (with some customisation) the same process again for the same types of matter? How do we develop our practices so that they are always on an upward iterative slope?

They may also ask: How can we drive out inefficiency and create efficiency, getting to the desired end result faster, but with less total input of time, energy, and legwork – while also reducing the total input costs and hopefully producing as good an output, or perhaps better and with even less risk for the end client? (And there’s plenty to consider here re. legal business models, but let’s do that on another day.)

But, how often is that ‘total work product’ approach actually happening, and happening via the prism of what genAI can do? This site knows plenty of people who do indeed think in those terms. But on the ground, amid the realpolitik of day-to-day delivery of legal services this perspective doesn’t always get more than a cursory hearing. The innovation signal gets lost amid the noise of the ever-churning billable hour machine’s constant rumble.

Small Steps Are Still Steps, But One Day You Have To Do More

There is nothing wrong in itself with small, one-off tasks, performed in isolation by individuals in a law firm, often in an ad hoc way. All such uses are valid, and this site will highlight many, and positively. Action, even if limited, is a lot better than stasis.

But, what happens if we stay there, in that random usage pattern? What happens if there is little methodical planning that deconstructs how things are made, practice by practice, office by office, team by team? Or when there is no real effort to consider the full extent and input depth of the legal production process for each type of client goal?

What happens if there is no concerted management-led effort from the top-down, as well as closely listened to feedback from the bottom-up, on how to structure the deployment of genAI across the firm, or inhouse team?

How can you change something as complex as a global law firm, with all its history, all its many partners, associates, support staff, client relationships, habits, culture(s), financial needs, and more, if there is no deliberate strategy in place that seeks to engage with the entire business structure and then drive change? We cannot expect the innovation team, or a well-funded vendor, or a handful of ‘tech champions’ in the partnership to do this on their own.

Breaking Stasis Needs Increasingly More Effort

So, small improvements are a welcome start, but if that’s it then they won’t change much systemically for lawyers. Crunching a few hours’ worth of ‘fiddly things’ with genAI a week is not going to move the dial. But, to reiterate, small steps are not a ‘bad thing’. We all have to start somewhere.

The issue is whether we move on from there, or not, or in fact have the impetus to break the inertia caused by the ever-growing complexity of legal work demands. I.e. when we are overwhelmed as an organisation we often tend to double-down on what we know has worked in the past, rather than stop to ask: ‘What if the answer is to change how we do things?’

The reality is that given the legal and regulatory environment will only continue to increase in complexity, that the largest corporates will continue to globalise and face new and even more challenging governance demands (including new AI regulations…) the need for something to give gets greater – just as the desire to hold on even more tightly to the past keeps us chained to well-worn, comfortable patterns.

Meanwhile, billable hour rates keep going up and the number of hours billed for tasks which have been done by the same firms 10,000 times before never seem to reduce, as if collectively the legal world has no memory and no material gains from technology have ever been achieved.

In short: more demand (over the long-term), more complexity (as a constant), and more costs for the commercial world (also a constant ­– at least under the standard model). And the whole thing just keeps rolling on. The standard model persists, and persists, and persists. Legal tech comes and goes. ALSPs also come and go. And still the standard model persists.

(Meanwhile, steadily rising legal costs mean that hopes for better access to justice remain continually frustrated. Although there is hope for real change here also via generative AI, but AL will tackle this in a separate piece given its importance.)

Two Conditions for Change (And One for Stasis)

So, what drives change? There are a few scenarios to consider. In fact, each of them deserves an article on its own, but for now here are three:

  • The New Tech Environment Scenario – whether lawyers like it or not, whether they oppose it or support it, change comes because the tech environment alters so systemically, so completely around them, that nothing can stand in its way. Email and the internet today may not seem very controversial, but they reshaped how lawyers work in multiple ways. Everyone adopted these tools and systems. It was impossible to operate otherwise. And so change just happened, just as winter turns into spring. (Although, in this case, email ironically both increased productivity and the speed of legal communications, but then made it easier to have endless redlines that in turn slowed deals down…..and let’s see where we get with genAI in that respect, i.e. that speed creates its own new challenges.)
  • The Innovators Scenario – in this one, some firms see opportunities to steal a march on their rivals, or perhaps rise up the rankings, via the use of new business models underpinned with new tech. Likewise, some clients’ legal teams – perhaps driven by central management ­– finally question the status quo (whether willingly or unwillingly) and become receptive to these innovators. This then bifurcates the law firm groups, as some rivals seek to compete, driving forward faster rates of change, while others dig in and double down with their confidence that their particular client roster won’t ever demand anything different. We saw this with the globalisation of law firms in the 1990s and 2000s – a form of strategic innovation, with the birth of many international and global firms which in effect moved in opposition to those seeking to be ‘independent’ and which mostly focused on their local market. In the end a new ‘steady state’ was found, although there was plenty of chaos in those intervening years. Independent firms also then created global referral networks, so they did globalise too in the end, just with a very different strategy. But, the size difference between these two groups is now so vast it’s hard to believe they were not that far apart back at the start of the 1990s. The gap now is mostly insurmountable. We may see the same happen in the next two decades with those law firms that really embrace genAI methodically, and those who just keep using it as ‘a little helper’.
  • The Convenience Conquers All Scenario – this is the stasis case and it also remains quite possible, i.e. where the new wave of AI languishes as a novelty that can shave an hour or two off a lawyer’s ever-increasing work volume, meaning that within 12 months its net value is effectively zero. In the stasis world lawyers don’t re-design how they work, so the benefits of automation (as that is what all software is ultimately doing in some form) are never realised. GenAI passes into the ‘legal tech fossil record’ and then one day ‘the next big thing comes along’ and it too, whatever it is, goes the same way….until one day either of the two other scenarios occur. That then finally raises the question: if this turns into another ‘legal tech fossil record’ moment, then when will real change come? It does seem almost impossible for the cumulative weight of tech environmental change not to eventually break the status quo. But, are we at ‘peak legal inefficiency’? This site would argue that this realisation has started to be felt by some.


It would be a bizarre twist of fate if the legal world finally got used to having an incredibly powerful new technology that really could change the entire sector for the better, yet never got beyond lawyers using genAI as just ‘a little helper’ or ‘a convenience’, which then had little material effect on our ‘standard model’.

Likewise, it could in many ways be the opening up of a new era for legal technology as it finally helps set us on the path to fully achieve the productive transformation it is capable of, but has not been allowed to do in the past.

And it’s worth saying the loss of lawyer roles is not the real threat here. The biggest risk is that the legal world fails to seize this opportunity.

Splitting the legal information atom with generative AI won’t blow up the legal world, it will give it a new lease of life. The future really is in our hands ­– we just have to decide to shape it.

By Richard Tromans, Founder, Artificial Lawyer, April 2024

[ Note: Robert Oppenheimer ‘father of the atomic bomb’ and Sam Altman of OpenAI, both share the same birthday.]