Government wants public input on use of AI

The province says it’s taking the lead when it comes to cybersecurity and artificial intelligence, but is it moving in the right direction?

Starting Aug. 10 in Brandon, Manitobans will get a chance to have their say on how the government should proceed when it comes to personal data security, the need to protect children from digital threats and the growing risks posed by AI.

“We want to hear from everyone because we want to make sure we’re getting this right,” Innovation and New Technology Minister Mike Moroz said Thursday. A second “public listening session” is set for Aug. 12 in Winnipeg.

Manitoba took the lead among Canadian provinces this spring when Premier Wab Kinew announced a plan to ban social media for youth under the age of 16. At the start of the 2024 school year, it banned cellphones in the classroom.

In June, Manitoba proclaimed the Public Sector Artificial Intelligence and Cybersecurity Governance Act. It says a public sector entity — the government, a government agency, health authority, university, school division or municipality — must comply with regulations governing the use of artificial intelligence systems and cybersecurity.

Regulations — which haven’t yet been written — could include having to provide information about the use of AI systems, the development of accountability frameworks and the reporting of cybersecurity incidents. The regulations could also set technical standards for the use of AI systems and cybersecurity. The minister may also issue directives about cybersecurity to specific public sector entities.

Moroz said he’s been hearing from a broad swath of Manitobans about the need for AI guardrails and cybersecurity since he was first appointed to the new cabinet post in late 2024. The public consultations invite more feedback, he said.

“We want to be thoughtful, we want to be deliberate and we’re not afraid to say, ‘This is the direction we’re going today based on the best possible information,’ knowing full well that the technology is changing so rapidly that six months from now or eight months from now, we may have to move some of the guardrails we put down because we discovered they were in the wrong spot,” the former school teacher said.

Moroz pointed to February’s mass shooting in Tumbler Ridge, B.C. as an example.

“Prior to Tumbler Ridge, nobody was talking about the downside of chatbots, then suddenly everyone was talking about it.”

The shooter’s ChatGPT account was flagged by internal systems and employees for generating scenarios involving gun violence months before eight people were killed.

“We could have laid down a guardrail around agentic AI, prior to that, only to discover that guardrail is in the wrong spot — we need to move it,” Moroz said.

Since Manitoba passed its AI legislation, the federal government introduced its own safe social media bill. It includes the Digital Safety Act and the Digital Safety Commission of Canada Act. The federal government said it is a framework to improve online safety and ensure that digital services are transparent and accountable for the risks their services create. It also creates a new Digital Safety Commission of Canada to enforce the Digital Safety Act and support victims of online harms.

Manitoba isn’t waiting for the federal government to take the lead, said Moroz.

“We’re going to continue to do what we think is necessary to protect Manitoba, to protect children.”

He’s consulting with federal ministers in charge of the bill and Manitoba’s feedback is welcomed, Moroz said.

“They need some provincial partners to make sure that what they’re attempting to do is done in a thoughtful, sensitive way. Having an active partner pushing in the same direction is important to them as well.”

Other provinces are starting to get on board with Manitoba to impose social media regulations on tech giants such as Meta and OpenAI, Moroz said. The minister said he’s asking states in the U.S. Midwest to join forces with their Canadian counterparts because there is strength in numbers.

“Meta is more interested in what Illinois says about some of these issues than what Manitoba says. The more partners, the more people we bring to the table, into the coalition, the stronger the voice we all have.”

Progressive Conservative innovation and new technology critic Josh Guenter said the public needs the province to be more transparent about its own handling of AI.

“Manitobans I have spoken with have raised concerns about the use of artificial intelligence and have a right to know how and when public institutions are using these new technologies,” Guenter said in a prepared statement.

“I am sure the minister will hear the same at his consultations; between this government and the federal Liberals, more must be done to ensure Manitobans are properly informed,” the member for Borderland said.

carol.sanders@freepress.mb.ca