Half of Canadian public sector workers who use artificial intelligence for work rely on publicly available AI tools, such as ChatGPT, according to a survey by consulting firm KPMG in Canada.

Using such AI tools exposes governments to potential risks, including data privacy and security breaches,

intellectual property theft and exposure to biased or inaccurate information that can lead to legal and ethical issues, the research found.

Ven Adamov, a partner and data analytics head at KPMG Canada, said that while there haven’t been any cases made public about sensitive government information breaches because of AI, the risk is still very real, as for any sort of confidential information.

“There is that risk (of exposing sensitive government data) and we need to educate people to avoid that from happening,” he said in an interview.

Just under a quarter of public sector organizations have adopted and implemented AI tools, but 48 per cent of their employees, ranging from client services agents to ministry workers to statisticians, use AI in their work anyway, the survey said.

Only 33 per cent of the workers are using private or proprietary AI tools provided by their employer, while 17 per cent use both public and private AI tools.

Of those surveyed, 32 per cent said their organization is currently experimenting with or piloting AI tools, while nine per cent said they intend to implement them. Eighteen per cent said their organization has not implemented AI and four per cent said they don’t plan to. The remaining 15 per cent did not know of any AI adoption plans.

“Because there’s a lot of nuances that are a layer on top of data governance, you need to pay attention to control and monitor these models as well,” said Adamov.

KPMG said AI can help the public service improve productivity, create new types of services, streamline or automate routine tasks and deliver services more efficiently and effectively to citizens and businesses.

However, Adamov said the public sector typically lags when it comes to innovation compared to the private sector, although governments are starting to adopt AI at scale and actually put policies and processes in place.

KPMG Canada’s research said AI adoption surged among private-sector employees after public AI tools, such as

OpenAI’s ChatGPT , became available in late 2022, after which companies moved rapidly to implement AI.

“This approach will not only protect public trust and data but also strike a balance between innovation and accountability,” KPMG Canada’s national leader of infrastructure, government and healthcare, Michael Klubal, said in a press release.

KPMG Canada surveyed 349 public servants in federal, provincial and municipal governments across the country from July 10 to Aug. 14.

The survey was conducted online, with 41 per cent of respondents working for the federal government, 29 per cent for provincial governments, 24 per cent for municipal or city governments and the remaining six per cent work at Crown corporations.

For those surveyed, the primary uses of AI tools include summarizing information, drafting or editing documents, providing preliminary research on topics and brainstorming and generating ideas, the survey said.

  • Carney urged to boost Canada’s digital sovereignty
  • Ottawa, Cohere to bring AI to the public sector

It said 37 per cent use AI a few times a week, 21 per cent use it a few times a month and 17 per cent use it daily.

A recent report by KPMG International and University of Melbourne found that Canada has the lowest levels of training, literacy and trust in AI systems in the world.