Canadian Lawyers and Generative AI: Some Suggested Starting Points for the Overwhelmed

How will generative AI impact the legal profession? This was the big question in late 2022 when ChatGPT was first released. How would this seemingly magical word machine collide or coexist with a profession that is, if not “essentially word merchants” (to use phrasing in a New York Times article), certainly deeply engaged with language?

Much has changed over the last two and a half years. To give just one illustrative example: fueled by hundreds of millions in investment, there are now over 500 legal tech tools that incorporate generative AI. The interest, money and work being poured into developing tailored generative AI tools for lawyers is truly remarkable.

However, notwithstanding all this development, the big question of 2022 is still the big question of 2025, in certain respects. How will generative AI impact the legal profession? This is still a very relevant question with a still unclear answer. The situation that we find ourselves in is highly dynamic. The technical capacity of generative AI technology is evolving rapidly, as are our understandings of this capacity. The situation is also characterized by a fair degree of uncertainty. Many tools are here, but how are they performing? There are growing and diverse efforts to “benchmark” and evaluate tools but these are still, in my view, in their relatively early days.

In such a dynamic and uncertain situation, some lawyers, in my experience, find themselves inclined to avoid generative AI. For these lawyers, this inclination is only further solidified when they see other lawyers becoming cautionary front-page news after having inadvertently misused generative AI and gotten in trouble for submitting fake cases to courts.

But, I don’t think turtling up is the way to go (probably not a surprise!). While we might not know where the generative AI plane will land for the legal profession, it has clearly taken off and there is good reason to believe that this technology will change, at least in some respects, the way that lawyers operate. Moreover, this technology is increasingly lurking all around us and getting built into existing tools that lawyers are already using. This sort of ubiquity makes it essential, in my view, for all lawyers to have some level of knowledge to engage with this technology properly. It isn’t just those who actively are choosing to buy the newest tools who are being exposed to generative AI. Add to this reality increasing client use of generative AI and need for lawyers to understand what their clients are doing and I think the case for engagement is undeniable. This is both a practical point and a professionalism one. Rules of professional conduct across the country are clear: to meet their competency obligations, Canadian lawyers need to “develop an understanding of, and ability to use, technology relevant to the nature and area of the lawyer’s practice and responsibilities” (see, Model Code, r. 3.1-2, Commentary [4A] and provincial and territorial equivalents). This doesn’t mean that lawyers must become computer scientists or tech wizards. But it does mean that a certain level of engagement is required.

So, if lawyers shouldn’t act like turtles, what should they do? There is no doubt lots that could be said about this from lots of different perspectives and levels of detail. I’ll offer a modest starting point here. In educating legal professionals about generative AI, I’ve recently taken to closing out my remarks with proposing a “3C AI Mindset.” My aim in doing so isn’t to offer a detailed “how to” on specific tools or uses, which would need to vary considerably between practice types and pre-existing levels of knowledge. Rather, my goal is to suggest a headspace from which lawyers can approach the AI moment that we find ourselves in.

First, it strikes me that it serves us well to approach the current moment with a curiousity mindset. Curiousity takes us away from feeling fearful or overwhelmed, which, as noted above, can be an easy and understandable default position in a world of fast-moving technology. If we are curious, we put ourselves in a learning mode. We open our minds rather than shut them down. I think curiousity is also valuable because it takes us away from a more passive or merely reactive mode. If we are curious, we can be curious about how we can be proactive and pursue approaches that are best for the legal profession and the justice system more broadly. If we are curious, we are more likely to be oriented to taking a critical approach to claims about what is inevitable and best for us. Curiousity opens up the range of possibilities.

My second suggestion is to approach where we are at with a confidence mindset. Being confident also takes us away from fear and overwhelm and helps put us in the driver’s seat. We should have confidence that we are a profession that has dealt with some big challenges before, including weathering a lot of technological change. Take something as commonplace now as email – we didn’t always have it and it was a big change for the profession when this technology was introduced. But, we found our way through (although I’m sure a lot of us also pine for the days when there wasn’t the degree of connectively that email enables). Here is also a good place to cue Ontario Attorney General Doug Downey’s remarks from 2020 where he observed that, in the response to the onset of the Covid-19 pandemic, our justice system “move[d] forward 25 years in 25 days.” The legal profession and the courts demonstrated remarkable, albeit imperfect, agility in that moment. To use a self-help mantra: we can do hard things!

Finally, adding a small dose of caution to all of this still seems healthy at this point. Again, we continue to be in a very dynamic time – opportunities and risks are still being sorted out, what is hype and what is reality is not always entirely clear, and the technology itself and its technical limits are far from settled. To be sure, we should be careful that our caution doesn’t spill over into fear and generate paralysis or encourage ignorance. But, also, it is okay for us to breathe and take a step back. Some aspects of this are uncharted territory and there is shifting ground.

Admittedly, this is all very general gesturing, but I do truly believe that approaching this moment with the curiousity and confidence that emboldens us to take action, while still being a little cautious and iterative, is a helpful mindset.

For those lawyers who are closer to the beginning of their generative AI journey and want more grounded ideas about where to begin than a law professor’s alliterative mindset suggestions, I’ll end with some brief, concrete recommendations for a generative AI “starter pack”:

1. Read law society and Canadian Bar Association guidance on generative AI.

One positive development over the past two and half years is the publication of broad guidance for lawyers on generative AI use (see my LinkedIn post here for a list). This guidance doesn’t answer all the questions that lawyers might have – it is general and high-level – but it is an accessible place for many lawyers to start and highlights some of the key issues to be thinking about.

2. Check out AI CPD events when you can.

One consequence of all the activity and interest in generative AI is that there is a steady stream of AI CPD offerings for lawyers. The price of these offerings varies, but there are a good number that are either free or low cost. For many lawyers, taking some sort of full or multi-day AI course is unrealistic but watching a webinar over lunch or listening to a podcast while you drive, ride or walk to work once a month is likely doable. Regular and repeated AI CPD engagement can help lawyers stay updated and facilitates hearing from multiple perspectives. Hearing from a variety of people on this topic is very helpful, in my experience. As AI expert Ethan Mollick observed earlier this year, the rapid technical developments in this area “means no one [including experts] has the whole picture now.” Different people have different pieces of the puzzle. AI commentators also speak, of course, from their own vantage points. People are understandably informed and motivated by their own worlds – generalizations are tricky, but it seems safe to observe that an AI vendor trying to sell a tool to lawyers is going to present a different perspective than an academic who, in turn, may be motivated by different concerns than a regulator.

3. Try out and experiment with tools in a safe way.

When it comes to AI, learning by doing is important. To be clear, “doing” here doesn’t necessarily mean using the latest and greatest generative AI legal tool. It can be simpler. To draw on Mollick’s insights again, he has observed “just use AI to do stuff that you do for work or fun, for about 10 hours, and you will figure out a remarkable amount.” This is simple, actionable advice for a beginner. For some, this might mean trying out general tools like ChatGPT. One other resource I’ve been recommending lately is the OBA’s AI Academy which was developed by OBA Innovator-in-Residence Colin Lachance and is an interactive and hands on platform that is free for OBA members.

AI can be a big, amorphous, scary, exciting, shifting and complex thing. But Canadian lawyers need not be overwhelmed. We can start small and be curious, confident and cautious on our individual paths forward.