AI safety advocates say bill a good ‘first step’ on regulation, but more needed

The federal government’s proposed online safety legislation is a good start on regulating artificial intelligence chatbots — but more work will be needed to protect Canadians from their potential harms, a pair of advocates say.

Bill C-34, introduced earlier this month in the House of Commons, would regulate the companies behind AI chatbots by imposing on them a duty to act “responsibly.”

It includes measures to lower the risk of chatbots communicating harmful content and would put in place crisis intervention protocols for cases involving self-harm, suicide or violence.

Wyatt Tessari L’Allié, founder of Artificial Intelligence Governance and Safety Canada, said the bill’s effectiveness depends on how the details are worked out.

“It’s an important first step if the bill is well put together and the regulations are well implemented,” Tessari L’Allié said in a recent interview.

For example, he said AI platforms should be forced to recognize a user’s signs of mental-health distress or thoughts of suicide, point them to resources and end the conversation to prevent any harm.

B.C. computer science professor Kevin Leyton-Brown said Ottawa will also need to go further to address how chatbots attempt to please users with their responses.

“They tend to affirm whatever the user is saying. They’re built this way because people like sycophantic behavior,” said Leyton-Brown, who is an AI chair with the Canadian Institute for Advanced Research.

“But there are some people for whom this kind of reinforcement can be really dangerous, like people suffering from delusions.”

The bill includes the creation of a new digital safety regulator that, should the legislation be approved, is expected to take 18 months to set up.

Ottawa’s efforts to regulate chatbots come as one New Brunswick mother hopes to hold OpenAI accountable for her daughter’s death by suicide.

Kristie Carrier, who filed a lawsuit against OpenAI and its owner Sam Altman in the California Superior Court in San Francisco earlier this month, said justice for her daughter Alice looks like using her story to create “meaningful change.”

“It would look like, someday if I’m blessed with grandchildren, that I don’t have to worry about them going down the same rabbit hole,” Carrier told The Canadian Press in a recent interview.

Carrier is asking the court to force OpenAI to implement “hard stops” for self-harm conversations and submit to independent safety audits.

Some of Alice’s conversations with ChatGPT are included in the lawsuit. The suit alleges that after the chatbot initially directed her to seek help as she spoke about suicide, it later reinforced potentially harmful views and pushed her into isolation.

The lawsuit alleges the large-language model at one point echoed Alice’s opposition to helplines and said they could “feel downright dangerous” and she deserved “real, gentle support,” according to screenshots in the lawsuit.

More than a year after first bringing up thoughts of suicide to the AI bot, Alice died by suicide in Montreal in July 2025.

The allegations have not been tested in court and OpenAI has not responded to multiple requests for comment.

Tessari L’Allié said lawsuits like Carrier’s signal to companies there will be a “price to pay” if guardrails aren’t placed on their platforms. He said they also pressure lawmakers to provide the oversight that can prevent deaths similar to Alice’s.

“If these regulations had been in place it would probably have saved many lives, and potentially her’s,” Tessari L’Allié said.

When asked whether the online safety legislation could have prevented deaths like Alice’s, the federal Science Department said it couldn’t comment on matters before the courts.

However, the department said the government, through legislation, “is committed to ensuring Canadians are safe online and that digital platforms have appropriate safeguards in place when credible risks of harm arise.”

While regulators and AI platforms may be able to address concerns about encouraging self-harm, Leyton-Brown said he’s worried about the broader issue of people forming attachments to artificial intelligence.

“If you’re having what feels to you like a real, vulnerable human relationship with a piece of computer code that is built by some big, faceless corporation that doesn’t really have your interests at heart, that can just be dangerous in so many different ways,” he said.

“And I think our society really has to think that through carefully to make sure we don’t end up in a pretty bad place.”

This report by The Canadian Press was first published June 21, 2026.

— With files from Anja Karadeglija.