An interview with Canadian technology lawyer Gillian Scott on how law firms can use AI to increase their efficiencies and what law firms need to know about various technology advices before deploying them in their practice. The article was first published in lexpert.ca.
- 1 An interview with Canadian technology lawyer Gillian Scott on how law firms can use AI to increase their efficiencies and what law firms need to know about various technology advices before deploying them in their practice. The article was first published in lexpert.ca.
- 1.1 What is legal automation and which services is it most used in? How widespread is it throughout the profession, and how long has it been in use?
- 1.2 What are the benefits to using automation technology as opposed to the use of human labour? Are there any reasons to be hesitant about implementing automated processes or products?
- 2 Benefits:
- 3 Reasons for hesitation:
- 3.1 Do you have any insights that you can offer law firms, practitioners, or in-house counsel that are looking to implement this technology?
- 3.2 Who are the actual “operators” of legal automation software once it has been adopted? Does a law firm or company need to have a dedicated team to its operation, or can it be outsourced to a third party? How much of successful adoption of automated processes is reliant upon close collaboration?
- 3.3 The legal profession is traditionally seen as one of the most conservative and least innovative occupations in the corporate world. What advice do you have for lawyers or firms that may believe that removing the human element to mundane tasks also runs the risk of overlooking AI-created errors and allowing them to multiply?
- 3.4 How do you see the automation in the legal sector evolving in the future? What is the next hurdle to wider adoption of it by lawyers, firms, and in-house counsel alike?
- 3.5 Do you have any other insights on helping clients navigate the use of legal automation services post/during COVID?
What is legal automation and which services is it most used in? How widespread is it throughout the profession, and how long has it been in use?
Legal automation is accomplished when we break down legal processes, or tasks typically performed by legal professionals, take some of those segments and embed them in technology. It is the crystallization or preservation of a piece of the “lawyer brain” in technology. The most prevalent legal automations involve complex high-volume work as well as common and repeatable tasks. We see them most frequently at work in high volume data collection, review, repeatable processes and document generation.
Lawyers and law firms have been using automation in various simplified forms in their legal work for decades. Think of a mail merge technology for addressing standard form letters, or automated tickler or reminder systems for deadlines in litigation. However, applications of legal automation have become increasingly more sophisticated, from culling large document review sets during document discovery in litigation, to pulling key terms from agreements in due diligence exercises. More recently, lawyers have been automating the generation of legal contracts and certain steps in a negotiation.
Automation is present in the practices of almost all lawyers and firms at a certain level when we are speaking about the most basic forms addressed above. For example, it would be rare at this point for a document review to be undertaken without e-discovery/document review software. However, the extent of the sophistication and penetration of automation across the industry varies widely between firms and practice areas.
What are the benefits to using automation technology as opposed to the use of human labour? Are there any reasons to be hesitant about implementing automated processes or products?
At its best, legal automation can remove repetitive tasks from a legal professional’s plate, freeing up their brains and time for complex work where they can truly add value. Legal automation can also assist legal service providers in delivering greater consistency, efficiency in their work product and client experience while better capturing data to inform better legal and business strategy and decision making. It is also used as a quality control measure to ensure that work product is accurate and reliable.
Reasons for hesitation:
It is often hard for those looking for legal automation solutions to know where to start. There are many legal tech providers in the market with myriad solutions on offer. Navigating which technologies are best suited to your needs, and which will incorporate best with your existing systems and processes can be a daunting task.
The initial investment costs of implementing technology can also be high, particularly in considering that many automating technologies will need to be particularly configured to your existing processes or systems.
Issues around data security are also critical for lawyers, law firms and legal departments with significant client confidentiality and privilege obligations.
The biggest hurdle, however, and the aspect that needs the most thought in a decision to adopt an automating technology is how the law firm or lawyer will promote adoption. Change management is difficult – especially for lawyers – who are by profile conservative and who are used to having done work in a particular way for a long time. They may see little reason to adapt or change their behaviours and have little time to engage with something they see as being outside of their core practice.
Do you have any insights that you can offer law firms, practitioners, or in-house counsel that are looking to implement this technology?
There is a perception or hope that after an outlay of certain funds for the purchase, license or build of a particular technology, that you will be “good to go” or that it will run itself: that legal automation will be a wave of a magic wand.
To the contrary, while great things can be accomplished with legal automation, there is a significant amount of advance work required for these technologies to be successfully implemented:
- With automation, you will still need a “lawyer brain”, to set up the various automation systems to do certain tasks. If the instructions or processes that are brought in at the front end during your build or configuration are wrong, you may end up replicating errors. Garbage in, garbage out. Further, as the law evolves, some of the logic built into your automations will also need to be updated to reflect those changes. Just as you can’t rely on a legal opinion that is five years old, you also can’t rely on an automated process that is five years old.
- You therefore need buy-in from those whose advice you wish to capture with the technology. This will take expertise and time. In most cases you will also need “translators” between the legal brain and the vendors or legal technologists implementing the processes in the software.
- Organizations should also have a very clear handle on their existing processes. If you automate a duplicative or flawed process you will be replicating or amplifying errors or inefficiencies.
- Perhaps most importantly, have a change management plan from day one. How will the technology be showcased to lawyers, to clients? How will you incentivize and drive internal and external adoption? How will the technology integrate with the lawyer or law firm’s existing systems or processes? How will you monitor the technology’s use and what will the KPIs be?
Businesses need to cover these bases before spending significant funds on a technology that could sit unused.
Who are the actual “operators” of legal automation software once it has been adopted? Does a law firm or company need to have a dedicated team to its operation, or can it be outsourced to a third party? How much of successful adoption of automated processes is reliant upon close collaboration?
Not to be too lawyerly, but it depends. Every organization will have to determine the right approach to implementation and that will depend on several considerations including:
- the workflow or document being automated
- the cost of the technology and its cost structure (i.e., is the automating tool proprietary to the firm, or is it licensed?), and the number of users it can realistically support.
As with all of the legal services we deliver and work we perform, I believe we owe it to our customers to think carefully about the most efficient way for us to provide the greatest value, and then set up a user model accordingly.
For example, our specialist lawyers and other legal professionals in our captive ALSPs, Osler Works – Disputes and Osler Works – Transactional, are the super users of most of our legal automation software and applications. Our lawyers, however, must have a strong understanding of what these applications are capable of, so that they can better instruct our specialist lawyers and communicate with clients about our capabilities.
In what I think is a very exciting development, we’ve opened some of our legal automation applications for direct client use or “self-serve.” With Osler Dash Select, the latest version of our automated franchise disclosure and agreement product, once set up on the Osler Dash platform, our franchisor clients can generate their own legal documentation.
The legal profession is traditionally seen as one of the most conservative and least innovative occupations in the corporate world. What advice do you have for lawyers or firms that may believe that removing the human element to mundane tasks also runs the risk of overlooking AI-created errors and allowing them to multiply?
First of all, AI is not a perfect salve to errors in legal work. AI-created errors can and do occur in instances where there is a complete reliance on the software. When people balk at using a document discovery software, it’s because they think the output should be perfect; and in some instances, it’s not. That’s OK. Neither AI nor humans are perfect, meaning that you need to have quality-control processes built in to support the software and ensure errors are caught. Automation doesn’t completely remove the necessity for a human element – in many instances, people are still involved, but what and how they work changes.
Second, automated tasks will be performed in a way that is as good as the thought and programming that went into the automation. If errors are made in the set-up, yes, those errors will be replicated. That is why at the front end you need to ensure that you obtain the right input, both from a technical and from a legal perspective. However, when an error does occur in an automated process, there should at least be a consistent audit trail as to how that error happened and where and when it was replicated. It can be fixed. Where the error is due to human discretion, or human inattention, it can be far harder to trace and undo. A good quality control process with checks is also essential.
Finally, of course, it is not every task, decision node or process that is ripe for automation. Where there is a strong element of legal judgment or discretion, tasks will require a lawyer or other legal professional. However, I do challenge lawyers to closely scrutinize their work, to understand it as a process, and as a series of decisions, and to really break down which points require actual lawyer brain, and which ones can be sorted and determined through logic programming. Conservatism on the part of lawyers when it comes to automation otherwise destroys value.
How do you see the automation in the legal sector evolving in the future? What is the next hurdle to wider adoption of it by lawyers, firms, and in-house counsel alike?
Automation in the legal profession will increase. Because of the cost of human capital and the drive for greater value in legal services, increased automation in the delivery of legal services is an inevitability. Where law firms and lawyers can use technology to improve their work product, their speed, their efficiency and their service delivery, it is imperative. This is a when and how, not an if problem.
Navigating the available solutions, understanding how those solutions fit within your organizations, and building an implementation and change management plan with buy-in from key stakeholders will be the greatest hurdles. Also note that in an increasingly automated world, the adoption of technology will be driven not only by our clients but by the court system, legislation, practice standards and professional guidelines.
COVID required many law firms and legal departments to rely on or adopt certain legal automation into their practices. E-signature technology is one piece that comes to mind. Businesses and firms that had balked at and fretted over electronic signature technology found themselves using it and seeing its efficiency and effectiveness. With the pandemic (hopefully!) receding, I doubt many of those businesses will revert to earlier wet signature practices.
COVID created a circumstance where firms and businesses had to adopt legal automation, but without the heavy lifting of change management and consensus around adoption. Necessity drove adaptation. Hopefully that experience, even though trial by fire, has shown law firms and legal departments that forward movement around technology can only benefit their businesses, and will perhaps allow them to embrace a less conservative attitude.
Gillian Scott is a partner specializing in innovation products at Osler Hoskin & Harcourt LLP in Toronto. Her role includes building out Osler’s overarching product strategy, as well as leading the market assessment, development, staffing, scaling, launch and sale of digitized product offerings. She brings 15 years as a commercial litigator and litigation partner at Osler and three years in Client Development to her role, giving her a unique understanding of client challenges and business objectives.