Years ago I thought the future of law would be Uber-like services that connect people with cheap lawyers in developing countries, or even semi-automated workflows run by less skilled people but acting according to playbooks. That didn't happen. Instead, ChatGPT happened, and it happened quickly. Now, millions of people in Canada are getting free legal advice by software, something that the Law Society of Ontario tried to stop for many years. Foreign companies just did it, and suddenly the genie is out of the bottle and the Law Society isn't trying to stop it. There is no factual difference between a software program that provides legal advice and a human (other than maybe differences in skill, but LLMs are getting better every month). In other words: the free practice of law is here.
Ontario law is currently being dished out in the form of ChatGPT dialogs, but also Gemini, and a host of other models. Originally they tried to stop people from being able to get legal advice but it seems those safeguards have been generally dropped and LLMs will happily opine on Ontario law now. Lawyers had no say in this, whether through their lobbying groups or professional organizations. It simply happened. And regulators haven't figured out even a position on this, let alone any way of trying to stop this.
It would actually be easy to stop ChatGPT from providing Ontario legal advice. I really think that it could be done, at least in the short term. The Law Society could simply insist on it, and OpenAI would stop. Same with Google and Microsoft. But maybe not some foreign LLM providers. In any event, that isn't what's happened, and it's unlikely to happen now that LLMs are becoming embedded into the practices of lawyers.
The era of the free practice of law is now upon us. There's a new world of legal advice that's suddenly accessible, and often quite good. While once it was primitive, often contract reviews by LLMs are functionally the same as what a skilled lawyer would provide, because the obvious points can now be detected by ChatGPT and other LLMs. The tech has caught up with at least a basic level of law, for some kinds of problems, and it's improving.
LLMs don't offer solicitor-client privilege, and they're much more limited and not as good as expert lawyers, but in 2025, the vast majority of people in Canada find lawyers to be impractical, largely due to cost. So LLMs it is. The public loses out on privilege (potentially), but gains speed, cost, and for most people it's the difference between something or nothing. This is the Uber For Law
future, but it's even cheaper or actually free. This is an amazing development that the legal community hasn't grappled with, other than in the context of fake citations submitted by litigators who are overworked/lazy/technologically inept.
This is probably the beginning of a new phase of human development in how people relate to the law. Suddenly, the extraordinarily large body of law that people are deemed to know might actually become knowable. Laws can be automatically integrated into tools because LLMs can actually do that. Not yet, but the path is set. Laws may become less optional, and people may not like the consequences of severely enforced law meted out by computers, but that's where things are going. Think AI-powered surveillance cameras that automatically set up prosecutions of people. And eventually, the prosecutions will be essentially a conveyor belt administered by humans, because this is the only cost-effective path in an LLM world. There are many ways things might go, and not all of them are positive, but there'll be more law that's actually experienced by people. The law has more meaning when it's accessible and understandable.
The free practice of law is now here. There was no vote. There won't be one. Let's see how it goes.