OpenAI reportedly violated Canadian privacy laws during the development of its widely-used ChatGPT tool, leading to the unauthorized collection and utilization of sensitive personal data. A collaborative inquiry by the federal privacy commissioner and provincial counterparts in Quebec, British Columbia, and Alberta unveiled these issues regarding ChatGPT, a conversational chatbot introduced in 2022 that mimics human-like responses to user inputs.
The investigation was initiated in 2023 after a complaint alleged OpenAI’s unlawful acquisition, use, and disclosure of personal information without consent. The examination revealed that the initial training of ChatGPT by OpenAI did not adhere to federal and provincial privacy regulations, as the company amassed significant amounts of personal data without adequate safeguards to prevent its exploitation for model training purposes.
The gathered information encompassed potentially sensitive details such as individuals’ health status, political affiliations, and even data concerning minors. Notably, many users were unaware that their data was being harvested and employed in ChatGPT’s training processes. Privacy Commissioner of Canada Philippe Dufresne expressed concerns about OpenAI’s hasty launch of ChatGPT without adequately addressing privacy issues, potentially exposing Canadians to risks like data breaches and discrimination.
In response to the investigation’s findings, OpenAI disputed the allegations, asserting compliance with privacy acts in most aspects. However, following the inquiry, OpenAI committed to enhancing its privacy protections and implementing additional measures to address the watchdogs’ concerns. The company detailed its data usage practices in model training, emphasizing the utilization of publicly accessible information and employing privacy filters to shield personal data within text inputs.
While OpenAI acknowledged the responsibility associated with users’ trust in ChatGPT for personal queries and tasks, Privacy Commissioner Dufresne stressed the necessity of updating Canada’s privacy laws to align with technological advancements like AI integration. The investigation coincided with the Tumbler Ridge mass shooting incident, prompting calls for regulatory measures targeting AI chatbots. Several legal actions have been filed against OpenAI and its co-founder Sam Altman in California, alleging negligence in handling the shooter’s ChatGPT account with disturbing content.
Dufresne advocated for a regulatory review to ensure privacy laws evolve in tandem with technological advancements, avoiding an immediate ban on chatbots and social media usage. Striking a balance between protecting children online and fostering digital literacy remains a crucial aspect of this evolving landscape.