Banks Cut Underperforming AI Agents
AI agents at major Canadian banks are facing a familiar human pressure: prove their worth or risk being replaced. The country’s Big Five banks are shifting from experimental AI projects to a more disciplined phase, where these systems must deliver tangible financial results.
According to consultants and businesses working with these banks, underperforming AI agents now face the robot equivalent of layoffs. The focus has moved from deploying technology for its own sake to ensuring it generates a clear return on investment.
Banks have invested heavily in AI. A tech research firm projects financial services companies worldwide will spend about $160 billion on AI tools in 2026, a figure expected to grow by 35.5 per cent over the next three years. While BMO, RBC and TD have set targets for AI returns, only TD has disclosed a dollar figure, reporting roughly $195 million in “AI value” in the first nine months of its fiscal year.
Banks Prioritize Results Over Experimentation
The era of expensive AI experimentation is over. Mitch Bouchard, general manager of Canada for ServiceNow, said the company is working with the Big Five to assess the business value of their AI agents. He noted that banks are now turning off projects that do not meet their standards.
“We’re seeing banks turn things off,” Bouchard said. “We’re seeing them bring more discipline to how they actually assess the value of AI.” This shift reflects a broader trend in the financial sector, where the focus is on outcomes, not just innovation for its own sake.
Success Depends on More Than Technology
Reuben Piryatinsky, a consultant who works with Canada’s big banks on AI, said the most successful implementations share common traits. Business teams, not tech teams, are held accountable for financial outcomes. The banks also create centralized AI standards and ensure their underlying processes are optimized before automating them.
“If your processes are broken, and you’re building AI agents to follow these processes, they’re just going to create even more of a mess,” Piryatinsky said. He emphasized that solid foundations are essential for effective AI deployment.
While underperforming AI agents may be cut, human jobs are not being eliminated. Alexandra Mousavizadeh, co-founder of Evident Insights, said banks are actually adding headcount. “Those 50 banks that we’re following are adding headcount rather than reducing it,” she said.
BMO Focuses on AI Outcomes
Greg Brown, BMO’s head of enterprise AI, said in a statement that progress will be measured by outcomes. “The measure of progress won’t be how many agents we deploy, but the outcomes they deliver,” he said. Other major banks did not provide comments by deadline.
