How generative AI transformed my practice as a junior associate – and what law firms owe the next generation
When I summered at a Bay Street firm, ChatGPT was the extent of most conversations about AI, and we certainly were not using it for legal work. By the time I returned to article, the AI-powered legal drafting tool Harvey had arrived, producing legal analysis that looked sophisticated enough that I could not yet confidently assess it. What once felt like a distant threat was suddenly part of my working day. If AI could perform work traditionally assigned to those entering the profession, what would become of articling classes and junior associate hiring?
Instead, I became an associate at another leading firm, working with the same calibre of lawyers and clients but in an office built around delivering work through technology. Learning to use emerging tools effectively is part of my job. Generative AI is central to that work today. The irony is not lost on me: I feared technology would narrow my career options, and I now occupy a role shaped by the opportunities it created.
What makes this a better job is not where I practise, but the kind of learning environment that technology has made possible. In my first year as an associate, I am part of a team where no one needs to pretend that AI does not exist. I can open a legal AI platform with a managing director behind me and feel no need to hide the screen. There is no contradiction between being a junior lawyer who still has much to learn and a lawyer who knows how to use technology to do the learning.
That freedom matters more than it may sound to someone with many years of experience. New lawyers are often handed discrete tasks without having the full history of a transaction, the client’s priorities, or even enough substantive knowledge to know why the assignment matters.
It is easy to call that poor delegation. Sometimes it is. But in large firms, people work across matters, deadlines, and time zones. The lawyer assigning the task may be in court, in the middle of a closing, or simply managing other competing demands. I never took the absence of an immediate explanation as evidence that my colleagues did not care. Often, I assumed the problem was me. Surely, I should understand this already. Surely, I should know what to ask.
An approved AI tool cannot tell me everything I need to know about a file, and its output still requires verification. But it can help me work through the first, unpolished questions: What does this term mean? Why might this clause be drafted this way? I can explore unfamiliar concepts, check the answers against authoritative materials, and approach a senior lawyer with a more focused question. Turning to AI for these answers is not an attempt to outsource judgment. It is an effort to develop enough understanding to exercise it. It makes better use of experienced lawyers’ time and expertise, while making the wait for their guidance feel less like a test of whether I belong.
And what happens when you remove some of the pressure to know everything before you’ve had the chance to learn it? You get someone more willing to ask questions and better prepared to arrive at a conversation with something more useful than an apology for not understanding.
Rethinking how we train junior lawyers
Imposter syndrome is not merely an unpleasant feature of early practice. For young lawyers who have invested tremendous time, effort, and money in their education and careers, the fear that they will never develop confidence, mastery, or excellence becomes a reason to leave the profession. If every day feels like evidence that you do not belong, it is difficult to see a long-term future in law or to stay long enough to experience the benefits of all your work.
Experienced lawyers are right to worry about what gets lost when junior lawyers no longer learn through the same repetitions. Those tasks taught discipline, pattern recognition, and an instinct for risk. An eloquent AI response can conceal a serious misunderstanding, especially from someone not yet equipped to spot it. Technological fluency is not legal judgment.
But if firms are already deciding what work should be done with AI, they must also decide how junior lawyers should be trained in an AI-enabled profession. The old training model cannot simply be removed without replacement. Sometimes, AI savings need to be reinvested in the very people whose work it is transforming. Developing the next generation will require greater intention, not less.
That adjustment will take time. But it may also create a more optimistic future for the profession. Imagine what might be possible if firms paired AI’s ability to ease some of the uncertainty of early practice with training that reflects how new lawyers now work. We might see more juniors develop the confidence and sense of belonging that make them want to stay. Clients build relationships with lawyers who grow alongside them. Firms retain institutional knowledge, build loyalty, and cultivate a stronger culture. Investing in how we train the next generation may also be an investment in keeping them.
I began articling, afraid that AI would eliminate my place in the profession. Instead, it helped create a role where my curiosity and willingness to embrace new technology are valued. I still want to learn from experienced lawyers, not just machines. But I now have a tool that helps me ask better questions and a workplace that encourages me to ask them. The problem is not that the next generation of lawyers will learn differently. It is that we might expect them to do so without rethinking how we teach them.
