While the goal of AI transparency is laudable, the federal government's consultation raises concerns about overreach and unclear expectations. Critics warn that vague standards could burden startups and small businesses with costly compliance without effectively improving accountability.
The consultation asks open-ended questions about 'explainability' and 'disclosure,' but lacks concrete definitions. This ambiguity may result in rules that are difficult to implement or that vary across sectors. Businesses fear that the voluntary code could become de facto regulation through public pressure, creating uncertainty.
Privacy advocates argue that the consultation does not go far enough; it lacks mechanisms for enforcement or penalties. Without binding obligations, companies may simply ignore the guidelines. Meanwhile, consumer groups worry that the process will be dominated by industry lobbyists, diluting protections.
The real risk is that the consultation delays meaningful regulation. The government has spent years developing AIDA, yet no law has passed. This latest consultation may be another stalling tactic, leaving Canadians unprotected from algorithmic harms.
To be effective, the government must move beyond consultations and enact clear, enforceable rules. Until then, transparency remains a goal rather than a guarantee.