Canada's Privacy Watchdog Secures Unprecedented Commitments from OpenAI

In a significant move, Canada’s privacy regulator has forced OpenAI to alter how its ChatGPT system operates for users in the country. The agreement, finalized last year, stems from a 2023 investigation that found the company violated federal privacy law.

The case began when a Canadian citizen discovered ChatGPT was inventing false and damaging biographical details about them. The Office of the Privacy Commissioner of Canada (OPC) determined OpenAI collected and used personal information without proper consent and failed to ensure the accuracy of the data its models produced.

To resolve the matter, OpenAI agreed to a series of binding changes. The company must now provide a way for Canadian users to challenge and correct false personal information generated by ChatGPT. It is also required to be more transparent about how it collects and uses data for training its AI models. Furthermore, OpenAI must implement measures to reduce the generation of such false information—a complex technical hurdle—and establish a formal complaint process.

The OPC will monitor the company's progress. If OpenAI fails to meet the terms, the regulator can seek court orders to enforce compliance.

This action places Canada among a group of nations, including Italy and South Korea, applying pressure on AI developers over data protection. It also highlights the limitations of Canada's decades-old privacy law in governing modern artificial intelligence, a technology its drafters never envisioned. The outcome sets a practical precedent for how regulators might hold AI companies accountable for the factual output of their systems.

Source: Webpronews