Telus and Fortanix: New Partnership Fixes AI’s Biggest Security Flaw

Telus and Fortanix have launched a new Confidential AI service that allows organizations to use sensitive information for AI training without it ever being unencrypted. The service runs on Nvidia hardware located at the Telus Sovereign AI Factory, ensuring all data stays within Canadian borders.

Traditionally, data must be decrypted to be processed by an AI, creating a security risk. This new technology, called Confidential Computing, keeps data encrypted even while the AI is actively working with it. The system uses what’s called a chain of trust that only unlocks data once the hardware is verified as secure.

“Telus has built Canada’s first fully Sovereign AI infrastructure so that researchers, startups, enterprises and public institutions no longer have to choose between protecting their data and unlocking the transformative benefits of AI,” said Hesham Fahmy, Chief Information Officer at Telus, in a statement. He added that this provides a “critical, final security layer to protect data while it is actively being processed.”

The announcement, made at the Nvidia GTC 2026 conference in San Jose, is aimed at highly regulated industries like healthcare, finance, and government. Before any data is accessed, the system performs a security check. If the hardware check fails, the encryption keys are not released, and the data remains locked.

“Trust is not implicit – it must be proven,” said Anuj Jaiswal, Chief Product & Strategy Officer at Fortanix. He noted that by using Telus’ infrastructure, companies can use AI on their most sensitive data “with cryptographic proof of protection, all on Canadian soil.”

The new service is now available through the Telus Sovereign AI Factory (in Rimouski, Quebec) using Nvidia’s infrastructure. A second Telus AI Factory is slated in Kamloops, BC, through expanding an existing data centre.