The federal heritage minister says the government is “seriously considering” whether to restrict young Canadians’ access to social media platforms and artificial intelligence chatbots — but hasn’t reached a decision yet. 

“We have some work to do, frankly, if we want to get it right,” Marc Miller said following a caucus meeting on Wednesday.

“Online harms don’t end as soon as you turn 15 or 16 or 17.”

The minister, whose portfolio includes online safety, was reacting to two non-binding resolutions Liberal Party members passed at their policy convention over the weekend in Montreal.

One urged the government to oblige social media companies to prevent users under the age of 16 from holding accounts. The wording is similar to a law passed by Australia late last year, which set a minimum age to possess an account on some of the most popular social media platforms.

The other carried resolution called for anyone under the age of 16 to be banned from accessing “all AI chatbots and other potentially harmful forms of AI interaction,” including OpenAI’s ChatGPT.

The government’s response to companies behind AI chatbots — tools that simulate human conversation — is being closely monitored in the wake of the Tumbler Ridge, B.C., mass shooting. OpenAI decided against flagging concerning messages the shooter had exchanged with ChatGPT months earlier.

Miller says a moratorium on social media and AI chatbots “could be an important layer” as the government looks to resurrect part of its online harms legislation, “but it has to be seen as that and not as the answer to everything.”

“The politics perhaps of it are convenient, but the policy has to be right,” said Miller.

He said legislation should focus on preventing online harms in the first place, “as opposed to suing afterwards or criminal liability.”

Artificial Intelligence Minister Evan Solomon said there are issues around age verification.

“How to do it? What’s the scope and scale?” he said on his way into the caucus meeting. 

Liberals reconvene panel of experts

Those types of questions are being put to a recently reconvened a panel of experts on the issue.

The group, made up of researchers, online child safety advocates and professors, has been asked to advise the government on how to approach the issue of online safety and regulate tech giants, including what to do about “the field of artificial intelligence, chatbots and AI companions, as well as other evolving trends related to online services,” said a Heritage Department statement.

The panel was previously brought together ahead of the former Bill C-63, known as the Online Harms Act, which died when former prime minister Justin Trudeau prorogued Parliament last year.

The divisive bill included Criminal Code and Canadian Human Rights Act amendments, targeting content used to bully a child or encourage a child to harm themselves, hate speech, content that incites violence or terrorism, content that sexualizes children or victims of sexual violence and sexual content that is posted without consent.

Conservative Leader Pierre Poilievre had accused the government of chilling free speech, saying bullying and other forms of online harms should be handled by police, not “pushed off to a new bureaucracy.”  

The Liberal government “cannot differentiate between hate speech and speech they hate,” he argued.  

Speaking Wednesday, Miller said previous iterations of the bill “overreached.” 

While the Liberals now have a majority thanks to five-floor crossers and three byelection wins, Miller said he will still look for support from the other parties on the anticipated new online harms bill.

“The initial sense I got from people that I did speak to on the other side of the House is that they would be interested in supporting the bill if it made sense for them,” he said.