Is Canada’s AI Minister just an industry spokesman now?

Open this photo in gallery:

Jake Pitre is a freelance writer and academic based in Montreal

Evan Solomon, Canada’s inaugural Minister of Artificial Intelligence, has been promising that the country will not miss out on AI’s economic benefits. He has been beating that message in with a stick.

That’s not a problem in itself. It’s Mr. Solomon’s job. He is barely two months into it, and his enthusiasm is commendable. His schedule has been chock-full this week with announcements, media interviews and appearances at tech events.

But that’s not Mr. Solomon’s only job. A minister in charge of an industry is not the head spokesperson for it. The minister also needs to be a supervisor, a neutral arbiter who can ensure the industry’s interests are not prioritized over the public’s. Unfortunately, it seems meaningful regulation of AI is to be a casualty of Ottawa’s myopic goal of boosting the industry.

This should disappoint Canadians precisely because this is the exact moment when urgent action must be taken to impose reasonable safeguards on this technology before it’s too late – and it will be too late very soon.

Mr. Solomon’s comments this week reinforce what he has long said, that Canada will move away from “overindexing on warnings and regulation” for AI, abandoning previously proposed rules while hyping up opportunities to commercialize AI and building many more data centres throughout the country.

Mr. Solomon’s positioning is not surprising. Much of this language echoes the Liberal Party’s Canada Strong plan, shared before the election. The plan demonstrated an aim to prioritize sovereignty and autonomy for Canada’s defence and security, in large part to rely less on the U.S., and all enabled by “the construction and development of AI infrastructure.”

Opinion: For the new Minister of AI, it’s easier to fail than to succeed

Like the party’s plan, Mr. Solomon‘s messaging fails to address the severe and looming threats posed by the industry on areas including privacy, security and its potential to reshape society.

He has made note of the previous government’s proposed legislation for protecting privacy, which targeted high-impact AI systems, but said the bill will not be reintroduced because “the climate has changed,” referring to the perceived failure of constraints imposed elsewhere in the world. This is a capitulation to the burgeoning but unrestrained AI industry, abandoning the idea that it requires sophisticated regulation at all.

This is disappointing because the new government has an opportunity to be a world leader when it comes to crafting reasonable guardrails on AI development at a time when so few governments seem willing or able to make meaningful moves as the technology evolves on a day-to-day basis.

Mr. Solomon has argued that regulation is not about finding a “saddle to throw on the bucking bronco called AI innovation … It is to make sure that the horse doesn’t kick people in the face. And we need to protect people’s data and their privacy.” That sounds sensible. But saying sensible-sounding words is easy. Without offering specifics, it amounts to letting AI run wild on our public and private institutions alike.

Mr. Solomon has noted that “the United States and China have no desire to buy into any constraint or regulation,” seemingly to use that as justification for neglecting regulations. That is wrong. The U.S. and China’s moves should not be an invitation to follow suit but to push back. Just because industry leaders are embracing the world of unchecked AI slop doesn’t mean Canadians must play by the same rules (or lack thereof).

Mr. Solomon’s position is even more condemnable when viewed in the context of current bilateral relations. If Canada, and the Liberal Party, plan to stand up to the U.S. and truly exert a persuasive and compelling sovereignty, accepting their policy mandates on such significant issues is not a viable first step. It’s cowardly.

Cohere to Canada’s tech industry: stay and grow at home

Then there is the matter of building new data centres and the immense power they consume. Emissions aside, Canadians could also see their electricity costs go up or become less reliable. AI data centres require more than 100 megawatts of electricity demand versus five to 10 MW for a regular data centre, so finding space within the electrical grid is easier said than done. Mr. Solomon has paid only lip service to this matter.

AI’s economic benefits themselves are far from proven. A recent report by Orgvue, the organizational planning platform, shows that more than half of businesses that rushed to impose AI simply made their employees redundant without clear gains in productivity. If a single-minded focus on these benefits is the way forward for Mr. Solomon and his ministry, it’s a shame, and a dangerous one.

Canada has an opportunity to be a world leader on imposing AI safeguards and making sure that our citizens enjoy its limited uses while being protected from its myriad threats to our economy and way of life. Unfortunately, that opportunity is quickly evaporating, and Mr. Solomon’s plan seems to be to sit idly by until it passes.