While the vast majority (80 per cent) of Canadian chief executive officers say their organizations are already in a rapid state of artificial intelligence (AI) adoption, only nine per cent of tech leaders believe their firms are fully prepared, according to new surveys from IBM.

One survey found the tech leaders (chief information officers and chief technology officers) are on track to deploy an average of 1,189 AI agents by 2027, representing a 36 per cent increase from today.

“There are significant desires; significant ambition at the CEO level, and they are convinced in using AI to realize value but there is a significant gap at the CTO and CIO level in establishing an infrastructure; establishing governance, and measuring the financial governance,” Raj Cherchattil, IBM Canada distinguished engineer and CTO for Canadian federal government, told PeopleNX in an interview.

Two separate IBM Institute for Business Value surveys heard from 2,000 CEOs in 33 geographies and 21 industries from February to April, and another 2,000 senior IT leaders in 33 geographies and 19 industries from January to April.

CEOs comfortable with AI 

The gap has exposed the varying levels of readiness and acceptance among the different levels in the C-suite: 90 per cent of CEOs surveyed are installing AI tools across all parts of the business structure, and another 65 per cent said they are comfortable using AI.

“The crux of the study gets the insights on what it would take to move from desires and ambitions into operationalizing AI, and potentially what are the control gaps,” Cherchattil said.

However, the adoption so far has only delivered measurable return-on-investment (ROI) in 43 per cent of cases over the past two years, the survey found.

“Financial governance and measuring the financial aspect of it is one of the largest gaping holes that we see,” Cherchattil said.

For most organizations, the debate around AI adoption is over. “It is a business imperative,” he said.

However, there are some steps that should be taken before widespread adoption is undertaken. “CEO-level members of organizations need to come together with the CIO and CTO to decide and define what problem are we solving, and what outcomes are we going to achieve that we can measure. Then the CIO and CTO organizations have to establish an IT strategy.”

In many cases, this step is not being done and that’s why there aren’t enough ROI success stories.

“It’s not small changes we’re talking about. You’re talking about major changes to ensure they have the right infrastructure adaptability,” Cherchattil said.

For CIOs and CTOs, they are feeling the gap most acutely: 68 per cent said in the survey they are accountable for AI systems that they don’t have full control over, and 73 per cent said the adoption is outrunning their abilities to manage it.

“There’s quite a bit of excitement whereby people are just creating what I call agent sprawls without the proper governance and management structure on top of it, and that usually creates confusion and possibly sub-par business results,” Cherchattil said.

"Significant gap" in skill levels

Another gap was exposed by the survey results: 89 per cent of CEOs believe the workforce has the skills required to fully implement and use AI. However, 53 per cent of workers need immediate upskilling today and 29 per cent of employees will need to reskill by 2028, according to the CEOs.

“Employees of these organizations absolutely need to go through the basic AI technology upskilling because a lot of the CEOs and C-level folks, right or wrong, they think that their employees are already skilled, which is not necessarily accurate,” Cherchattil said.

“There is a significant gap in the skills.”

But beside fully educating employees around AI usage, the training also needs to happen in the C-suite, according to Cherchattil.

“In my opinion, they need to understand and be reskilled on the gaps in the ambition versus reality on the ground in terms of their organizational capabilities. Where do they have weaknesses, and what actions do they need to take?”

The survey found that 68 per cent of tech leaders cite this lack of training as a major barrier preventing AI adoption.