Of Canadian public servants polled, 48 percent utilized AI for their jobs, 50 percent of which resorted to publicly available tools, resulting in risks like data privacy and security breaches, intellectual property theft, exposure to false or biased information, and legal and ethical issues, KPMG LLP said.
KPMG in Canada surveyed 349 public servants in Canada’s federal, provincial, and municipal governments. The results revealed that 33 percent of those using AI at work had access to private or proprietary tools from their employers, while 17 percent utilized public and private AI tools alike.
KPMG explained that survey respondents mainly used AI to summarize information, draft or edit documents, conduct preliminary research, brainstorm, and generate ideas. Regarding frequency of use:
- 37 percent used AI a few times a week
- 21 percent used it a few times a month
- 17 percent used it daily or on every project
- 15 percent cited an ad hoc basis
- Six percent tried it but barely used it
- Four percent used it once every few weeks
According to those KPMG polled, among their organizations, 22 percent adopted AI tools, 32 percent were piloting or experimenting with such tools, 18 percent did not implement AI, nine percent intended to implement AI, four percent had no such plans, and the remaining 15 percent of respondents had no idea about their organization’s AI adoption plans.
“With public sector employees already adopting AI tools to carry out their job responsibilities, public sector organizations must accelerate their deployment of formal AI adoption policies,” said Michael Klubal – KPMG in Canada’s national leader of infrastructure, government, and healthcare – in the press release.
Among respondents to KPMG’s survey, 52 percent did not utilize AI tools in their jobs. Of that figure:
- 30 percent saw no advantage to using AI at work
- 30 percent believed they lacked the knowledge or training to utilize these tools
- 17 percent did not trust AI
- 12 percent reported that their employers prohibited AI use
“Our survey highlights the need for strong governance, oversight, and thorough training to roll out responsible and secure AI adoption and embed best practices,” Klubal said. “This approach will not only protect public trust and data but also strike a balance between innovation and accountability.”
KPMG noted that 36 percent considered AI “extremely” and “very” important for their organizations to be more productive and efficient, 42 percent deemed it “somewhat” important, and 22 percent dubbed it unimportant.
Challenges
KPMG’s research identified the following obstacles to adopting AI:
- job security fears
- ethical concerns
- privacy requirements
- security and compliance issues
- deepfakes and other fake content
- disinformation and misinformation
- low trust in data quality
- inadequate readiness of risk management processes and governance
- insufficient standards and regulations
According to KPMG, 80 percent of respondents reported that their organizations lacked data professionals guiding the ethical use of data and establishing the appropriate data governance frameworks, while only 55 percent said their organizations had clearly defined data roles and responsibilities.
“It takes a team that includes data scientists, data engineers, software developers, and ethical AI specialists to build and stand-up AI systems, not to mention critical thinking and creativity,” said Ven Adamov, a partner and data analytics leader in KPMG in Canada’s risk consulting practice, in the press release.
KPMG’s survey also covered possible deficiencies in terms of data readiness. The survey showed that, among respondents:
- 39 percent did not have or know about a clearly communicated data policy
- 49 percent did not have or know of a clearly communicated data strategy
- 60 percent had not undergone data literacy training
KPMG shared that 69 percent of respondents belonging to organizations that had adopted AI believed that AI anxiety was real, while 77 percent of all respondents said their organizations feared losing control over the AI system and its results.
KPMG stated that 85 percent of respondents rated AI literacy among public sector employees as low, while 82 percent rated AI fluency as low.
In its press release, KPMG said 85 percent thought Canadians did not fully comprehend or feel confident in governmental plans to use and develop AI.
KPMG’s research looked into perceptions regarding the protection of Canadian data. Among respondents:
- 94 percent believed that data storage should be within Canada
- 87 percent wanted to incentivize private industries and large cloud service providers to keep sensitive or personal citizen data in the country
- 86 percent thought that storing data outside of Canada might erode public trust
“When data is moved to the cloud, security becomes a shared responsibility with third-party cloud service providers,” Adamov said. “Governments do not have full sovereignty over their data when stored in the cloud, and the data could be subject to foreign laws and be disclosed to another government.”