OpenAI found to have violated Canadian privacy law in ChatGPT training

Canada’s federal privacy commissioner and counterparts in Quebec, British Columbia and Alberta have found that OpenAI did not follow Canadian privacy laws when it trained ChatGPT. The joint investigation said the company collected and used sensitive personal information without adequate safeguards or clear user consent. The probe began in 2023 after a complaint alleging unlawful collection, use and disclosure of personal data. Regulators said the issue involved largescale data gathering that could include health details, political views and information about children, while many users were not aware their data could be used in model training. OpenAI disputed parts of the findings but has since taken steps to improve privacy protections and said it will add further measures. Canadian privacy commissioner Philippe Dufresne said the case shows current laws need to be modernized to better address AI tools and protect Canadians’ privacy rights.