Canadian firms race ahead on AI, but readiness lags

Kyndryl has released new Canadian findings showing that businesses are moving artificial intelligence into everyday operations. The data points to a widening gap between adoption and readiness to scale AI securely.

Its 2026 Readiness Report found that 80% of Canadian organisations are deploying AI in workflows beyond pilot projects, while 66% reported a positive return on investment. At the same time, readiness to manage external risks fell to 22% from 31% a year earlier.

The figures suggest many companies have moved from testing AI tools to embedding them in routine business activity, but have not kept pace with governance, infrastructure and workforce changes. Only 25% of Canadian organisations have formally defined accountability for AI-related risks, decisions and outcomes.

Canadian executives also reported weaknesses in the underlying technology needed to support broader use. Some 62% said significant re-architecture is required before their technology foundation can support AI at scale, and 47% said technical debt is holding them back.

Workforce preparation emerged as another obstacle. Just 26% of respondents completely agreed that roles have been redesigned around AI, while 36% said their workforce is ready to use it effectively.

The report also found that strategy remains unsettled for many organisations. Some 43% of Canadian leaders said they arrived at their current AI strategy by accident, suggesting adoption has often moved faster than formal planning.

The Canadian findings reflect a broader global pattern in Kyndryl's survey of 3,700 senior business leaders across 21 countries. Globally, 64% of leaders reported positive returns on AI investments, up 10 percentage points from 2025, while average readiness for external risks fell to 24%, down seven points.

Security concerns remained the top external risk in the global data. Only 26% of firms said they were completely ready for cyberattacks, an 11-point drop from the previous year, even though three in four leaders said AI helps their organisations identify threats faster.

Infrastructure strain

The report draws a distinction between modernising IT systems and making them fit for large-scale AI use. While 71% of global leaders said their IT estate is mostly or fully modernised, 66% of Chief Information Officers said their infrastructure still needs significant re-architecture before it can support AI at scale.

That issue appears particularly relevant in Canada, where the share of leaders citing a need for re-architecture was lower than the global figure but still substantial. The findings suggest that investment in digital systems has not necessarily produced an environment ready for the computing, data and governance demands of wider AI deployment.

People and control

The global findings also showed uneven progress in adapting organisations for AI. Just 27% had formally updated roles and responsibilities around AI, and 29% had fully implemented change management for AI adoption. Organisations that had done both were more likely to report readiness and positive returns, according to Kyndryl.

Nearly half of global respondents, 48%, said they had substantially changed their AI strategies over the past year. The same dataset showed that about a quarter had seen an AI investment fail, while 21% had cancelled a major AI initiative.

Martin Schroeter, Chairman and Chief Executive Officer at Kyndryl, said the challenge for organisations has changed as AI has moved further into core operations. "As AI creates significant opportunity, organizations are now focusing on how to make early returns sustainable," he said.

He added: "This year's Readiness Report shows a new kind of readiness gap: AI ambitions are moving faster than the systems needed to support them, and the bar for readiness keeps moving as AI becomes more integrated in enterprises. Our data shows that focusing on people, technology and governance can help organizations win with AI and keep the ability to adapt as opportunities evolve."

The Canadian results come as policymakers place greater emphasis on the secure use of artificial intelligence across the economy and public sector. Against that backdrop, the findings indicate that many organisations have moved past the pilot stage but still lack the structures needed to manage AI consistently as it becomes part of day-to-day decision-making and operations.