Corporate Canada is witnessing an unexpected reversal: a growing number of employers are rehiring staff they'd previously replaced with AI. The reason is straightforward - the real cost of AI tools has caught up with, and in many cases surpassed, the price of full-time employees. A Robert Half survey of 1,365 professional-services hiring managers found that 34% of firms that cut staff after adopting AI have already reinstated those positions or similar roles, in a pattern the firm calls "AI correction hires."
The hidden price of AI tokens
Early AI adopters were drawn in by flat-rate subscriptions and monthly fees from companies like OpenAI and Anthropic. Those models are shifting to usage-based billing, measured in units called tokens. For many businesses, the bill is landing far above what a human worker would cost.
"This is the first time ever that I can remember that technology costs the same as people," Arvind Jain, CEO of Glean, told CNBC. Nvidia's Vice-President of Applied Deep Learning, Bryan Catanzaro, told Axios that the "cost of compute is far beyond the costs of the employees" on his team. According to Zylo's 2026 SaaS Management Index, organizations spent an average of US$1.2 million on AI-first applications in 2025 - more than double the previous year.
OpenAI CEO Sam Altman acknowledged the backlash directly. "People are really saying … 'My company spent my entire 2026 budget in Q1. Can you make this more efficient?'" Altman said. "That went from, at the beginning of this year, an issue that never came up … to, all of a sudden, a huge issue."
Microsoft cuts staff, then cuts Claude
Microsoft's recent moves illustrate the tension. The company pushed US$80 billion into AI data centres and let go of 15,000 people worldwide - roughly 4% of its workforce - in 2025. This April, it offered voluntary buyouts to senior-level employees whose age plus years of service totalled 70 or more.
Soon afterward, Microsoft cancelled its developers' Claude Code licenses after about six months of use and redirected engineers to GitHub Copilot CLI, a switch widely attributed to the unexpectedly high cost of agentic AI tool usage. The company is still embedding AI deeply across its products, but it is increasingly building proprietary software to avoid per-token charges.
The 'AI boomerang' hits Canadian desks
Across the market, the boomerang is well documented. Klarna, the global fintech firm, dropped 700 customer service representative contracts in favour of an AI chatbot. The move saved US$10 million initially, but customer dissatisfaction forced the company to rehire its human workforce.
Lisa Cohen, associate professor of organizational behaviour at McGill University's Desautels Faculty of Management, explains why most AI systems only replace slices of work, not whole jobs. "You can't just take tasks in and out without changing the connected systems," she said. Gartner research predicts that globally, half of businesses that laid off customer service or operations workers will rehire them by 2027.
What the data shows about AI and jobs in Canada
Statistics Canada reports that the percentage of Canadian businesses using AI to produce goods or deliver services more than doubled from 6.1% in 2023-2024 to 12.2% in 2024-2025. Yet among AI adopters, only 6.3% reported a decrease in employment because of AI - a figure that held steady even as adoption surged.
Roughly 60% of Canadian workers, about 12 million people, are in jobs considered highly exposed to AI-driven transformation. About half of those workers are in roles where AI is more likely to augment their work than eliminate it outright, according to StatCan.
Cost reality check for HR leaders
Robert Half's data carries a warning for employers planning cuts driven by AI. Three-quarters of employers who reduced headcount to save money found that the costs of rehiring, retraining, and lost institutional knowledge erased those savings.
For HR managers steering the change, the ground rules are shifting. Pilot AI in limited workflows and track real token costs against the fully loaded cost of labour - including rehiring, benefits, and productivity losses from turnover - before removing any roles. Redesign job descriptions to blend AI literacy with the complex judgment and communication tasks that tools cannot handle. HR professionals can explore dedicated resources on AI for Human Resources to guide workforce planning and role redesign decisions.
Why this matters for HR professionals
The "AI boomerang" is not a sign that automation is failing; it is a signal that the financial picture is more nuanced than many boardrooms assumed. HR leaders are uniquely positioned to insist on total-cost analyses that include token consumption, training, and rehiring expenses before headcount decisions are made. Building internal capability in AI workforce analytics and people strategy - for instance through an AI Learning Path for HR Managers - can equip teams to advise the business on where augmentation works and where replacement will backfire. The organizations that navigate this shift without costly reversals will be those that treat AI adoption as a workforce design challenge, not a simple subtraction of payroll.
Your membership also unlocks:
