Why is making the best use of lawyers time an issue? Lawyers are busy people…

Legal AI: A human centred approach to process improvement required
We hear a lot about the potential for Artificial Intelligence (AI) to augment the work that lawyers do.
But how to determine which tasks are best done solely by AI and which tasks are best done by human lawyers with support from AI?
While this may seem to be new and unchartered territory, it’s not.
The practice of process improvement and subsequent application of process automation tools has a relatively long history in legal services. For example, over 25 years ago I worked on mapping legal processes and then automating them using workflow software.
Using a traditional approach to process automation, we can understand what tasks lawyers perform, how they perform them and how their performance may be enhanced (in the present context by AI).
This is classic process improvement, and it provides a good starting point for considering the augmentation of lawyers’ work by AI.
However, to ensure we have the balance right between work done by AI and human lawyers we should develop a more nuanced approach: a human centred approach to process improvement.
In this article I will explain what the human centred approach to process improvement is and provide some principles about how to apply it.
First, let’s briefly remind ourselves what traditional legal process improvement is and how it is applied in practice.
What is legal process improvement?
The International Institute of Legal Project Management (IILPM) defines legal process improvement as:
A structured methodology for optimizing legal and business processes, so that legal professionals can deliver high-quality, cost-effective services in less time and with less effort.
The structured methodology most often chosen as a starting point for legal process improvement is Lean / Six Sigma (i.e. combining core principles of Lean and Six Sigma).
Lean
Broadly, Lean techniques help ensure teams are ‘doing the right things’.
Ideally, legal teams should only work with processes, and tasks within them, which add value to clients. Any tasks which do not clearly add value (and are not required by law) can be eliminated from the process under consideration.
How to assess whether tasks are ‘adding value’? A quick rule of thumb is to ask, ‘if clients knew we were doing this task in this way, would they be willing to pay us for doing it?’.
If the answer is ‘no’ then we need to see whether this task can be removed from the process and / or whether task performance can be improved.
Six Sigma
Six Sigma techniques are more concerned with ‘doing things right’. Making the remaining tasks which add value the most efficient and effective they can be.
One of the core concepts found in Six Sigma is that of reducing variation of task execution and quality of output.
If three different people were to perform the same task and left to their own devices, chances are the way each perform that task will vary and quality of output will vary.
What we should be aiming for is high quality output every time, regardless as to who is performing the task. We can achieve this by reducing variation of task execution by providing things such as process checklists, approved document templates, staff training and, of course, automating tasks whenever possible.
Automating the work of legal service teams
For some, the ideal is for computer software to execute legal tasks without needing intervention of humans at all (whether they be lawyers or anyone else in legal service teams).
Practice areas such as debt recovery have long been targets for process automation and it is unsurprising that the first AI law firm (i.e. the legal work of the firm being done entirely by legal AI) regulated by the Solicitors Regulation Authority is focused on debt recovery.
Practice areas like this are ideal candidates for traditional Lean / Six Sigma analysis as a prelude to process automation.
Augmenting the work of lawyers
Let us now consider the goal of augmenting some of the work which lawyers do.
To do this we need to understand not only the tasks that lawyers and their teams perform but how they perform them, especially remembering the uniquely human parts of task execution.
Assuming we understand the processes and tasks making up a legal workflow (following on from a Lean / Six Sigma approach) we now need to understand what skills lawyers and their teams have and how they apply them.
Some core legal service skills are listed below, including:
- Analysis, Research & Problem-Solving
- Legal reasoning: Interpreting laws, statutes, and case precedents to form well-reasoned arguments.
- Legal research: looking up any case law, statues and precedent material relevant to the matter at hand.
- Critical thinking: Evaluating legal issues from multiple perspectives to craft strategies.
- Issue spotting: Identifying potential legal risks and implications within contracts, litigation, or corporate matters.
- Communication & Advocacy
- Client communication: Translating complex legal jargon into actionable advice.
- Persuasive writing: Covering both the technical, such as drafting effective contracts and, following on from the previous point, making sure client communications are clear, concise and understandable.
- Business & Strategy Acumen
- Project management: Ensuring the effective delivery of their services to clients
- Negotiation skills: Balancing legal and commercial interests in deals and settlements.
- Financial literacy: Understanding billing structures, risk management, and cost-effective legal solutions.
- Ethical & Professional Judgment
- Adherence to legal ethics: Maintaining confidentiality, avoiding conflicts of interest, and ensuring fairness.
- Decision-making under pressure: Balancing legal obligations with business and ethical considerations.
- Emotional intelligence: Managing client relationships with empathy and strategic insight.
How can we assess which combination of tasks and skills can be augmented at the most appropriate point during the legal workflow process and which ensure consistency of results?
Service Design Principles
Service Design emphasises a qualitative approach to understanding and improving processes, with the focus being on client need and delivering value to clients.
Perhaps the most important principle of Service Design is that services should be designed and delivered in collaboration with all stakeholders, both internal and external.
The focus is on interactions between all stakeholders engaged in delivering or experiencing the service under review.
Noticing interactions between stakeholders is important, as it’s here where uniquely human skills, such as those listed above, most often come into play.
Human Centred Process Improvement
I think of Human Centred Process Improvement as the application of Lean / Six Sigma and Service Design principles.
Taking our cue from Service Design, human-centred process improvement should place the needs, desires, and experiences of individuals — clients, employees, and other stakeholders — at the centre of our thinking when assessing which parts of legal service work can be best augmented by AI.
The idea is to move beyond streamlining and automating processes, useful though that is.
This should lead us towards automating many standard tasks while providing more time and space to apply the skills that human legal service teams bring.
And by the way, one human legal skill not listed in those above is a relatively new, but increasingly important one: checking and verifying the output of AI’s work product.
How to make a start with Human Centred Process Improvement?
Because legal processes cannot be seen (in contrast to manufacturing processes) most legal process improvement work is usually done via workshops.
Workshop participants (the process improvement team) work in the process under consideration and ideally their process improvement efforts should be guided by a Legal Process Improvement Professional.
Legal Process Improvement Professionals should be familiar with all the well-known process improvement methods and techniques and be able to adapt them to circumstances as required.
They must also have well developed facilitation skills, working with all stakeholders to understand the interaction between task execution and the application of legal skills.
Thereafter facilitators need to foster collaboration from all stakeholders to seek and develop improvements in the workflow – in this case, by deploying AI appropriately.
All this is easier said than done, but personal experience shows it can be done.
Process improvement methodologies provide tried and tested means for understanding and improving processes of all kinds.
Structured process improvement methods have a lot in common with each other, such as they
- focus on the need to identify what clients value and will pay for
- are workshop driven (especially in service environments)
- make a lot of use of graphical charts of various kinds (such as value stream maps or client journey maps) and
- use various brainstorming techniques during workshops to help generate potential solutions.
Human Centred Process Improvement Skills
Many legal service organisations are still at the early stages of formally employing AI technology but given the pace at which the technology is developing, they need to move faster.
One huge constraint they face is the lack of process improvement skills available in their organisation. Even in legal service organisations which have invested in developing their process improvement capability, there are not enough process improvement professionals to meet demand.
This demand for legal process improvement skills will probably increase during the next year or so as legal service organisations race to catch up with AI deployment.
If you would like to find out more about how you or your colleagues can acquire legal process improvement skills and become a Legal Process Improvement Professional, please see my legal process improvement course pages and visit the International Institute of Legal Project Management, which has further information about legal process improvement training and certification.