Yes. Thank you very much for that question.
The prior Liberal bill was introduced without consultation and without input from the general public. It had a structure that had concentrated all the regulatory authority within just one ministry—industry. It didn't recognize that, in fact, AI is like electricity, microchips and other technologies that are pervasive throughout the economy.
As such, it tried to regulate the technology as opposed to regulating the uses. If you're regulating uses, Health Canada might deal with AI and health, OSFI might deal with AI and financial institutions and there may be certain cases where there wasn't one particular department that would have been useful, but that wasn't the approach that they took. They tried to model it on just an outline of a bill with everything being left to regulation, which in my view undermined the importance of Parliament making a decision about AI policy. It would have been all in the hands of the people who were promoting regulation. In my view, this was not the appropriate model.
I had looked to the U.K. at the time, which had a decentralized model with a centralized coordination structure. I think that's a much better model for AI legislation. Bill C-27 created uncertainty because nobody knew what exactly was going to be regulated.