Canadian privacy authorities conclude that OpenAI broke federal and provincial privacy laws in the process of developing ChatGPT, according to reports. The findings say that personal information from Canadians was included in data used to train the company’s AI model. Regulators also cite shortcomings in how OpenAI handled collection and safeguards for that information, and they raise concerns about whether the company had valid consent for the use of personal data in training.
The investigations focus on the period when OpenAI built its first ChatGPT model, and they characterize the privacy issues as tied to how data was gathered and protected during development. One report adds that, since the probe began, OpenAI has addressed “major concerns,” suggesting the company has made changes in response to the regulators’ findings.
The decision increases regulatory pressure on AI developers operating in Canada, as privacy watchdogs examine how companies use large datasets and what obligations apply when personal information may be included in training material.
