error
Subscriptions are not available for this site while you are logged into your current account.
close
Skip to main content

      Artificial intelligence is spreading rapidly across Canadian organizations, yet measurable returns remain scarce. According to KPMG's November 2025 survey of Canadian business leaders, 93 per cent of business leaders report using or piloting AI technologies, up from 61 per cent last year, but only two per cent say they are realizing measurable returns on investment. This article explores why adoption alone is not enough to generate tangible business impact, and why further investment in the people, process, and governance layers is essential for translating AI adoption into sustained value.

      AI adoption is rising but organizational gaps persist

      AI adoption is widespread, but integration is not. Many organizations have introduced AI through pilots or point solutions, yet only 31 per cent have embedded generative AI across core operations and workflows.

      On the employee side, AI adoption remains divided between early adopters and a large minority that is still on the sidelines. Data from the 2025 KPMG Generative AI Adoption Index shows that 51 per cent of Canadian employees now use generative AI at work, up from 46 per cent in 2024. Among those users, most report tangible benefits with 79 per cent stating that generative AI has improved their productivity and more than half save between one and five hours of work per week.

      While employee AI adoption has grown, the pace is slowing. After rising sharply from 22 per cent in 2023 to 46 per cent in 2024, the share of employees using generative AI increased by only five points in 2025 to reach 51 per cent. Nearly half of employees (49 per cent) still aren't using AI at work, showing that momentum is tapering off. 

      There may be several contributing factors:  
      • Low trust in AI

        In KPMG's global study with the University of Melbourne, Canada ranked 42nd out of 47 countries in trust in AI systems.

      • Limited AI literacy and training

        The same study found Canada ranked 44th in AI literacy and training, indicating a significant skills gap.

      • Insufficient organizational guidance

        Only 29 per cent of Canadian employees say their employer has a comprehensive policy outlining acceptable use cases, up from 18 per cent last year.

      • Restricted access to generative AI tools

        7 per cent of employees report that their employer still prohibits the use of generative AI tools, although this has fallen from 16 per cent in 2024.

      Combined, the data reveals a consistent pattern: AI technology adoption is increasing, but the human and organizational capabilities needed to generate measurable outcomes are still catching up. Across industry, organizations have invested heavily in technology, while the corresponding people and process pillars – including workforce literacy, change management, and workflow redesign – remain underdeveloped.

      Alongside these gaps, governance has become an essential fourth layer, providing the accountability, value measurement, and steering mechanisms needed to navigate responsible AI adoption. Together, these capability gaps form the core of Canada's AI readiness challenge: without balanced investment in people, process, and governance, technology adoption alone is not likely to translate into tangible business impact.

      KPMG logo
      Stephanie Terrill
      Stephanie Terrill

      Canadian Managing Partner, Digital and Transformation and National Leader, Management Consulting

      KPMG Canada

      Employees need more targeted training to build AI literacy

      The KPMG Generative AI Adoption Index 2025 found that among employees that use generative AI at work, 83 per cent believe they need to improve their skills to use AI effectively. However, fewer than half (48 per cent) of them say their organization currently offers sufficient training on how to use AI effectively in their role.

      In contrast, 82 per cent of executives said their organization provides AI training, up from 74 per cent in 2024. This suggests that training is increasingly available, but many employees still do not feel it is helping them use AI to be more productive.

      Closing this literacy gap requires more tailored, role-specific learning. As the Vector Institute notes, different stakeholder groups have distinct training needs:  

      Stakeholder Group
      Training Focus
      AI Builders – Teams who design, develop, and maintain AI systems.Creating reliable, scalable, and secure AI tools.
      AI Users – Employees who apply AI tools and capabilities as part of their work.Understanding opportunities, limitations, and safe practices for AI.
      Control Functions – Groups such as risk, legal, and compliance that oversee governance, policy, and risk management.Becoming effective counterparts to AI Builders and AI Users to challenge their assumptions and reveal gaps.
      Support Functions – Functions like IT and HR that enable the organization through technology, people, and operational support.Securing AI-ready resources for the organization.
      Senior Management – Leaders who set the strategic direction for AI and allocate resources.Identifying trends and anticipating market and environmental shifts.  

      Limited trust and unclear governance may be hindering AI adoption

      Responsible AI use is growing, but trust in the technology remains limited. Findings from the KPMG Generative AI Adoption Index 2025 show that 58 per cent of employees now always check AI outputs for accuracy, up from 51 per cent in 2024, and fewer are entering private or proprietary company information into prompts.

      Despite this progress in responsible AI practices, trust in AI remains low. Fifty-six per cent of employees are very or extremely concerned about hallucinations (false or misleading outputs) and say this limits their use of AI at work. Broader public sentiment reflects similar concerns. In KPMG's global study with the University of Melbourne, nearly half (46 per cent) of Canadians believe the risks of AI outweigh the benefits, with top worries including cybersecurity risk (87 per cent) and loss of privacy or intellectual property (86 per cent).

      Employee awareness of AI governance and guardrails is also limited. Only 29 per cent of employees say their employer has a comprehensive AI policy, while 42 per cent are unsure whether one exists. This uncertainty can leave many employees unsure how to use AI responsibly or what their employer expects of them.

      As the Vector Institute highlights, existing governance frameworks were not designed with AI in mind. Because the AI landscape evolves faster than policy can, many teams must make day-to-day decisions without clear, use-case-specific guidance. To close this gap, Vector recommends establishing clear, principle-based guidance aligned with corporate values, and training employees to apply those principles in real-life situations. This should then be complemented with the longer-term work required to update governance frameworks to make them AI-ready.

      Limited AI integration may be holding back business ROI

      Although 93 per cent of organizations report using AI in some capacity, most remain far from full deployment. Only 31 per cent have moved beyond proofs of concept and pilots to implement AI across core operations, while the rest are still in early stages: 32 per cent with partial deployment into select workflows and 20 per cent still testing or piloting.

      Organizations continue to report that generative AI has yet to produce measurable business returns, a finding that reflects the current state of deployment. With most implementations still in pilot or partially deployed phases, the conditions for meaningful value capture are not yet in place. Only two per cent of respondents said their organizations are seeing a return on their generative AI investment. Among those reporting a return, more than half (57 per cent) described it as modest (between five and 20 per cent) while nearly one third (31 per cent) were unable to quantify it at all.

      Behind these modest returns lies a common challenge: many organizations still face foundational barriers that prevent AI from scaling across the enterprise. Legacy systems, siloed data, and unclear success metrics often prevent AI from being embedded throughout the organization.  

      Low AI integration may also reflect a lack of clear strategy for selecting the right use cases. Companies should not underestimate the effort it requires to select, refine, and mature their use cases, ensuring that they create both tangible business value and personal value for employees. When AI tools address real pain points and make day-to-day work easier, employees are far more likely to adopt, adapt, and sustain use.

      A further contributor to low ROI may be the tendency to layer AI on top of existing processes rather than redesigning workflows to take full advantage of the technology. Integrating AI into current workflows can create incremental improvements, but the greatest value comes from AI-optimized processes where tasks, decisions, and systems and redesigned to leverage AI's full capabilities. 

      Acting boldly to realize the promise of AI and unlock its full value

      Only a small proportion of Canadian organizations are generating measurable returns from their AI investments today. That trend is understandable as new technologies take time to be adopted and show results. Canada, however, is facing near-term pressures on competitiveness and productivity, and waiting years for AI investments to pay off is no longer viable. To stay competitive, organizations must first strengthen AI foundations. This includes building workforce literacy and establishing robust governance mechanisms that ensure responsible use. These capabilities are essential to creating the trust and confidence required for responsible deployment.  

      To capture the full value of AI, organizations also need clear ROI frameworks that measure not only financial outcomes but also strategic, operational, and capacity gains. When paired with strong governance and accountability, these frameworks help translate AI ambition into measurable impact.

      Canadian business leaders are no longer debating whether to invest in AI; they are focused on how to adopt and scale it responsibly and effectively. Success will depend not just on the technology, but on the people and processes that make AI valuable. When these foundations are in place, organizations are far better positioned to accelerate deployment, embed AI into core functions, and achieve measurable value creation.

      How we can help

      ac_unit

      Discover how our end-to-end AI consulting solutions can accelerate AI with confidence.
      Learn more
      public

      Empower your AI strategy with ethical and responsible AI solutions.
      Learn more

      Connect with us

      KPMG. Make the Difference.

      We’re here to help your organization thrive.

      Connect with us
      building