Most of the world today uses AI chatbots on a daily basis for just about everything.
They can be incredibly useful and provide quick answers to questions that might require minutes or hours of research.
I am a self-described luddite. Even so, I regularly use AI chatbots.
At home, AI can help with the most basic tasks. My recent queries have included:
- Is tomorrow a good day to hang clothes on the line?
- What’s the best recipe for roast beef sandwiches?
- What’s the value of my car?
- Why won’t my chainsaw start?
At work, we have specialized AI tools that can assist with everything from legal research to summarizing medical records to preparing chronologies.
Both at home and at work, AI is a great starting point. Unfortunately, it’s well known to hallucinate and make mistakes. A healthy dose of common sense, critical thinking, and verification must be applied to every answer it returns.
For important queries, one is better off just consulting an expert.
Should You Trust AI for Legal Advice?
When it comes to the practice of law, I’ll let you in on a secret: your lawyer is probably using AI in 2026. Even before AI, your lawyer probably googled some things. There are lots of reliable legal sources on the internet which can help answer complex questions.
So why not just use AI yourself to answer legal questions? Simple – there is lots of unreliable information out there too. Your lawyer is almost certainly better than you are at:
- Choosing the correct AI chatbot and feeding it the correct prompts and information to generate a correct response; and
- Critically thinking about the response (and what is missing from it) and verifying the sources to which AI refers.
AI’s History of Bad Advice: When Chatbots Get it Wrong
There is no shortage of examples of AI’s bad advice. It has become somewhat of an internet meme (examples here, here, and here).
In a recent incident, a B.C. teen was hiking and had to be rescued from a dangerous trail. While he consulted AI for assistance in planning the hike, he says AI was not to blame for his wrong turn. Thankfully he was rescued and unharmed in this hiking incident.
Most of AI’s bad advice is obviously wrong, or relatively harmless, but what if AI gives bad advice and things go really wrong? Can you sue? Depending on the circumstances, yes.
Because the prevalence of AI chatbots is relatively new, lawsuits against AI companies are new as well.
The most notorious are probably the current lawsuits against OpenAI for its failure to report to authorities some of the prompts submitted by the shooter in the tragic Tumbler Ridge shootings.
Other cases in Canada and the United States include:
- Moffatt v. Air Canada (British Columbia, 2024): A customer relied on incorrect bereavement-fare information from Air Canada’s website chatbot. The tribunal held Air Canada responsible for the chatbot’s misleading information and awarded $812.02 for negligent misrepresentation. (Moffatt v. Air Canada, 2024 BCCRT 149 (CanLII))
- Nelson v. OpenAI (California, filed 2025): The parents of a man who died from an overdose sued OpenAI. They allege ChatGPT gave dangerous medical advice, including recommending a lethal combination of Xanax and kratom, and failed to advise him to seek medical attention. (Parents Sue OpenAI After ChatGPT Medical Advice is Blamed …)
- Setzer v. Character Technologies (Florida, filed 2024): A wrongful-death action alleges that interactions with Character.AI chatbots contributed to the suicide of a 14-year-old boy. (Judge Denies Artificial Intelligence Chatbot First Amendment …)
- Nippon Life v. OpenAI (Illinois, filed 2026) : A lawsuit alleging that ChatGPT tortiously interfered with a settlement by providing legal advice and drafting documents that caused a person to breach a settlement agreement. (Designed to Cross: Why Nippon Life v. OpenAI Is a Product Liability Case)
I have not been able to find a case yet in Canada where a bad-advice negligence lawsuit was actually tried and the Court rendered a decision. The closest thing I could find is the Moffat v. Air Canada case.
First Principles: What Do You Need for a Negligence Lawsuit in Ontario?
First off, you typically only sue legal persons (humans, corporations). Chatbots aren’t legal persons, so you have to sue the creator.
Also – most of the AI creators likely have forum selection clauses you must agree to as part of the terms and conditions for using the Chatbot. This means your ability to sue them in Ontario might be limited, although forum selection clauses aren’t always enforceable.
For a successful negligence lawsuit in Ontario, you need the following:
- The Defendant owed the Plaintiff a duty of care
- The Defendant breached that duty (or fell below the standard of care required)
- The Defendant’s breach caused the Plaintiff harm
While there’s no established Canadian case law yet delineating what duty AI creators owe to Chatbot users, I would hazard a guess that AI giants do owe quite a significant duty to users not to cause them harm.
The standard will be harder to define. For example – does it meet the standard if it gives bad legal information with the caveat “you should consult a lawyer”? Or – if it tells you how to perform electrical work and you get electrocuted – should you have been smart enough not to use AI for such a task?
Only time will tell how these cases play out.
Will We See More AI Negligence Lawsuits in Canada?
Don’t get me wrong – AI is an excellent tool. But as we move forward, I predict we’ll see more and more lawsuits for its bad advice.
If you or a loved one has been injured in Ontario and have questions about whether someone else may be legally responsible, contact Bergeron Clifford Injury Lawyers for a free consultation. We serve clients in Ottawa, Kingston, and across Eastern Ontario.
