Ontario's Workplace Safety and Insurance Board (WSIB) laid off 566 employees this month, and workers and their union say the agency's push to implement AI tools played a central role in the cuts. The affected staff include case managers and customer service representatives who say they were asked to train the very systems that replaced them.
The Ontario Compensation Employees Union (OCEU), which represents WSIB workers, announced the job cuts earlier this month. Union president Harry Goslin told CBC News the layoffs are "about automating the work."
Workers say AI training preceded layoffs
Brittnee Gauthier, a post-adjustment case manager in Sudbury who lost her job, said WSIB made "a huge push" internally to implement a new AI claim summary tool that reviews and summarizes documents. She said she was processing more than 200 claims when she received her layoff notice.
"I can only deduce that the future will be AI for this company," Gauthier said. "I'm sure AI can help. We all were excited about this tool that was getting launched - until we found out it was making us all redundant. So, I just worry about who's next."
Sanjeev Kapoor, a WSIB customer service representative in Sudbury who was also laid off, said he and others were told to train an AI system by checking whether it gave correct answers and verifying policy numbers and dates. "We were kind of training the AI to take over our own jobs," Kapoor said, adding that the AI had often provided incorrect policy numbers. Goslin told CBC News other employees have raised similar concerns about the accuracy and reliability of the agency's AI.
Contract details and leadership ties
WSIB records show it awarded a $7.3 million contract to Owl.co for a "claims summary tool subscription" in February. In October, WSIB CEO Jeff Lang is scheduled to speak at a Las Vegas conference alongside Owl.co CEO Sean Merat about how "WSIB is deploying AI at extraordinary scale."
WSIB told CBC News the layoffs were a "difficult business decision" and that it is "always looking at ways to modernize our business and deliver the best support for people while delivering value for the premiums Ontario businesses pay." The agency said the claim summary tool "organizes information and documents so team members can understand a claim more quickly."
The agency did not provide answers on the record to questions about the accuracy of its AI tools or whether it plans to use other AI programs. The union has filed grievances about "the withholding of information concerning technological changes that are not being disclosed," Goslin said, and has also filed freedom of information requests.
Impact on injured workers
Employees who were laid off said they were notified on a brief group call with hundreds of other people. Gauthier said she is now rushing to process her claims "because people are just going to be falling off." She believes "injured workers are going to feel the brunt of this."
Goslin said legislation requires that decisions on claims be made on their merits in each case. "AI is only as good as you program it and the bumpers that you put in place," he said, raising concerns that services to claimants will be diminished.
WSIB told CBC that it "continues to be here to help people after a workplace injury or illness all across Ontario. How people access our services has not changed." WSIB is funded by individual companies and not by tax dollars, but it is a provincial Crown corporation reporting to - but operating at "arm's length" from - Ontario's Ministry of Labour, Training and Skills Development.
Why this matters for insurance professionals
This case shows what happens when AI tools are deployed before their accuracy is proven and before workers understand how automation will affect their roles. For insurance professionals, the pattern is familiar: claims summarization, document review, and policy verification are exactly the tasks AI vendors are pitching to insurers across the industry. The WSIB situation suggests those tools can reduce headcount quickly - and that errors in policy numbers and dates are a real risk when AI handles regulated claims work. If your employer introduces AI tools, ask directly how accuracy is measured, what oversight exists, and whether training the system is part of your job description. Those questions are easier to answer before layoffs are announced.
Your membership also unlocks:
