Nouveautés

The Law Society of Alberta has already drafted your firm's AI rules. You just have to adapt them.

It is a free, four-page guide for your lawyers and staff, with the firm's AI use rules already written. Download it, fill in your firm's details, and adjust the rules to fit your practice.

If you have not got to this yet, you are in good company. In Clio's 2025 Legal Trends Report, 53% of U.S. legal professionals said their firm has no AI policy, or that they do not know of one.

While you adapt it, add two questions for every AI tool your people use: 1. Is our client information used to train it? 2. Can privileged documents stay out of it?

Use AI on the whole case file, without putting privilege at risk.

Most lawyers think it's one or the other. Sort out the privileged documents by hand before any tool sees the file, or don't use the tool.

In Juris you mark a document privileged once. From then on, Sage leaves it out of every answer, every summary and every draft it builds for that case, and works through everything else.

The speed on what you can share. The documents you can't stay yours. And every privilege mark is logged with who made it, so you can show it was handled.

You know your file too well. That is the problem.

By the fourth read you are not reading any more, you are recognising. The date you glanced past on the second read is the one you will glance past on the tenth. The other side finds it on their first.

A twenty-lawyer firm hands the file to someone who has never seen it. Solo practice has nobody to hand it to.

Juris is that person. Point it at the documents and it comes back with strengths, red flags, the contradictions across your own record, what is missing and who to get it from, and the opportunities you have not taken.

AI is right there and free until you hit a limit, so of course people are using it to pull a case or check a cite. Nobody is being reckless. That is just a Tuesday afternoon with a filing due.

The trouble is these tools will hand you a case that does not exist, and they will also hand you a real case that says something it never said, and both of those look exactly like good research right up until someone checks.

Lawyers have been sanctioned for it. Costs out of pocket, referrals to the law society, and a name sitting in a published judgment that does not go anywhere.

Nobody sets out to file a case that is not real. It just needs one busy week.

A lawyer filed two cases that ChatGPT invented. The court accepted she had not meant to mislead anyone. She still paid the other side's costs out of her own pocket, and the judgment carries her name.

The part worth reading twice: in the New York matter, counsel did check. He went back to the model and asked whether the cases were real, and it told him they were.

A model asked to confirm its own output will usually confirm it. Agreeing is the most plausible continuation of that conversation. It has no separate memory of what it made up.

The check takes about ninety seconds, and it is written up here.