OpenAI Faces Class Action Lawsuit for Alleged Unauthorized Personal Data Collection
A law firm based in California has recently taken legal action against OpenAI. The Clarkson Law Firm filed a class-action lawsuit claiming that OpenAI has been collecting personal data without authorization to train its AI language models, specifically ChatGPT and DALL-E.
The lawsuit alleges that OpenAI obtained private information, including personally identifiable data, from a significant number of internet users without their consent or knowledge. This complaint has been filed in the Northern District of California court.
Allegations of Stolen Personal Data and Lack of Consent
The allegations in the lawsuit suggest that OpenAI gathered an extensive volume of data, estimated to be around 300 billion words, from the internet, which included personal information and content from prominent social media platforms such as Twitter and Reddit. Clarkson Law Firm contends that OpenAI carried out this data collection covertly and without fulfilling the necessary requirements of registering as a data broker, as mandated by relevant laws.
The law firm highlights the absence of informed consent or awareness on the part of the individuals whose data is claimed to have been utilized for training ChatGPT and DALL-E.
OpenAI's Controversial Data Collection Practices
OpenAI has been at the center of controversy regarding its data collection practices. Previously, there was no explicit option for users to opt out of sharing their conversations and personal information with OpenAI for model training purposes. Furthermore, ChatGPT faced a ban in Italy under Europe's General Data Protection Regulation (GDPR) due to concerns over inadequate protection of user data, particularly that of minors. This class-action lawsuit not only highlights OpenAI's opaque privacy policies but also focuses on data that was scraped from the web without being explicitly intended for use in training AI models.
Lawsuit Highlights Privacy Policy Concerns and Data Use
The lawsuit raises significant concerns about OpenAI's privacy policies, especially regarding existing users. It alleges that OpenAI has profited from the data it collected without compensating the source, leveraging billion-dollar investments from Microsoft and revenue generated through ChatGPT Plus subscriptions. The complaint specifically accuses OpenAI of violating privacy, negligence in safeguarding personal data, and larceny by illegally obtaining massive amounts of personal information to train its AI models.
The Counts in the Complaint Against OpenAI
The complaint lists a total of 15 counts against OpenAI, encompassing various allegations related to privacy violations, negligence, and illegal acquisition of personal data. The counts highlight OpenAI's alleged misuse of publicly available datasets, such as Common Crawl, Wikipedia, and Reddit, without obtaining proper permission or consent from users. Although personal information shared on social media platforms and other public domains may be accessible, the lawsuit argues that using such data outside of the intended platform could constitute a violation of privacy.
Legal Distinctions and the Debate Over Data Use in the U.S.
In Europe, legal distinctions exist between public domain and free-to-use data, thanks to the GDPR. However, in the United States, this issue remains a topic of debate. Nader Henein, a privacy research VP at Gartner, supports the sentiment of the lawsuit, emphasizing the importance of individuals having control over how their data is used, even when it is publicly available. Nevertheless, Henein expresses uncertainty regarding the stance the U.S. legal system will take on this matter.
The outcome of the lawsuit against OpenAI remains uncertain. As the legal proceedings unfold, it will be interesting to see how the court addresses the allegations of stolen data and the violation of privacy. This case could have significant implications for the data collection practices of AI companies and the protection of the personal information of individuals in the digital age.