Summary
Recently, OHRC Director of Policy, Education, Monitoring and Outreach, Juliette Nicolet, and IPC Commissioner of Ontario, Patricia Kosseim, joined Radio-Canada « Le matin du Nord » for an interview on Principles for the Responsible Use of AI in Ontario.
02:35:09. Éric Robitaille: We can follow you on TikTok, but especially listen to you on Fridays on Matin du Nord. A big thank you for being here. Have a nice weekend! Thank you, see you soon. You're listening to Le Matin du Nord. We're now going to talk about security regarding the advancement of technologies and the spectacular developments in artificial intelligence, which has been quickly disrupting our lives in recent years. More and more tools are being created, and even more companies are opting to invest in artificial intelligence, which is really everywhere. And like all technology, it has many advantages. It’s incredibly useful, but it's also full of equally immense dangers, particularly for privacy and human rights. This is why Ontario now has principles to guide the responsible adoption of artificial intelligence systems, principles that were adopted jointly by the two organizations that delegated the guests I'm now presenting to you. First, Patricia Kosseim, Ontario's Information and Privacy Commissioner. Hello Ms. Kosseim,
02:36:13. Patricia Kosseim: Hello, Eric.
02:36:14. Éric Robitaille : And also with us, Juliette Nicolet, Director of Policy, Education, Monitoring and Outreach at the Ontario Human Rights Commission. Hello Ms. Nicolet. First, just to fully understand which sectors, which organizations, are going to be the subject of the guidelines you've established, Patricia, tell us more about the mission, the mandate you've been given, and in what context artificial intelligence will be regulated.
02:36:42. Patricia Kosseim: It's precisely in the context of the public sector in Ontario.
02:36:47. Éric Robitaille: So, anything related to government departments or even agencies hired by the departments?
02:36:51. Patricia Kosseim: Exactly, exactly.
02:36:53. Éric Robitaille: And Juliette, what exactly was your role in these protocols you established?
02:36:59. Juliette Nicolet: Well, we're very happy and satisfied with the relationship we have with the IPC. We've worked in detail and in collaboration to develop the principles, and we believe they are principles that will above all help to advance public trust in the use of artificial intelligence in the public sphere.
02:37:22. Éric Robitaille: Patricia, tell us about the dangers and risks you've identified, before we talk about solutions and guidelines.
02:37:30. Patricia Kosseim: Of course, as you mentioned, it brings a lot of potential and benefits, but also risks. In particular, risks to human rights, risks to equity, risks of discrimination, as well as risks to the security of personal information and the privacy of Ontarians. So that's precisely what we're trying to minimize with these guidelines.
02:37:57. Éric Robitaille: So, Juliette, tell us about what you've established as a guideline.
02:38:02. Juliette Nicolet: Well, the principles are clear. We believe that the use of artificial intelligence, the systems themselves, must be valid and reliable, they must be secure, they must respect privacy, they must protect human rights, they must be transparent and accountable. And in doing so, I think what the Commissioner said is very true. We see, we envision this great innovation that is transforming lives and society. It's inevitable. And we simply have to set strict guidelines to ensure that privacy and human rights are protected as we move forward.
02:38:43. Éric Robitaille: So, I turn back to the commissioner, Patricia Kosseim. We're well aware of the principles, but in concrete terms, what kind of gesture, adaptation or tool will this provide to the Department and to organizations to avoid slip-ups and danger?
02:38:58. Patricia Kosseim: The guidelines, I think, offer parameters for institutions looking to develop or use this technology. I think we've developed guidelines that are practical and that disclose our expectations as regulators of private and human rights in a very transparent manner.
02:39:23. Éric Robitaille: So, artificial intelligence systems must be valid, reliable, secure, privacy-friendly, must protect human rights, must be transparent, accountable. Juliette, do we have models elsewhere? You've established the policy for Ontario, but have we taken inspiration from formulas that work, or don't work elsewhere and tried to correct them?
02:39:42. Juliette Nicolet: Obviously, we looked at a wide range of resources, laws and frameworks on artificial intelligence at the provincial, national and international levels. This is a very new context. The need for supervision is evident everywhere. On the other hand, the Ontario Human Rights Commission along with the Law Commission of Ontario, we launched together in 2024 an Artificial Intelligence Impact Assessment Tool that is designed to help assess, to mitigate the human rights impacts of AI systems in a very wide range of applications, both in the public and private sectors.
02:40:30. Éric Robitaille: I understand, Juliette, that we've emphasized the fact that you can't leave artificial intelligence to work on its own, so you always need a framework of human beings who know the tool, who can ensure that the guidelines are respected.
02:40:44. Juliette Nicolet: Yes, whether it's a human being or a machine, one cannot be exempt of responsibility with regard to obligations in terms of the right to privacy, data protection, human rights. And so, the job at hand is to proactively identify and eliminate these problems in the design and implementation of AI systems based on grounds protected by the Human Rights Code as well as data protection and protection of privacy.
02:41:14. Éric Robitaille: Juliette Nicolet, you're Director of Policy, Education, Monitoring and Outreach at the Ontario Human Rights Commission. Patricia Kosseim, you are the Information and Privacy Commissioner of Ontario. A huge thank you to both of you for giving us your time this morning. Thank you.
02:41:28. Patricia Kosseim: Have a nice day. Thank you.
02:41:30. Éric Robitialle: You're listening to Le Matin du Nord.