OpenAI privacy violations in Canada have been confirmed by federal and provincial privacy commissioners in BC, Alberta, and Quebec. Carmi Levy says the response was a blog post and that is not going to change much.

The findings are specific. OpenAI collected health information, political views, and data on children from the open internet without asking consent. It did not vet what it scraped, meaning false information entered the training data. And when Canadians found errors about themselves, there was no clear mechanism to correct them. Carmi Levy tested this himself when ChatGPT launched in November 2022: he asked it what it knew about him, found a lot, and found some of it wrong.

The deeper problem is the legislation. Canada’s Privacy Act comes from 1983 and the commissioners can flag violations but currently lack the tools to enforce meaningful change. OpenAI is preparing to go public, which Carmi Levy says tells you exactly how much internal cleanup is happening right now.

Topics: OpenAI privacy Canada, Canadian privacy commissioner, data collection AI, ChatGPT violations, AI privacy laws

GUEST: Carmi Levy

Originally aired on2026-05-08