OpenAI and Microsoft are facing yet another legal battle, as they have been named in a lawsuit filed in a US District Court. The complaint alleges that OpenAI’s AI model, ChatGPT, has been using personal data to train its models without obtaining proper permission. This comes just a month after OpenAI was sued for defamation for generating false statements about a radio host.
The recent lawsuit is focused on the alleged non-consensual use of personal data. Though the plaintiffs in the case have remained anonymous, the complaint seeks class-action status and mentions potential damages of $3 billion due to the involvement of millions of class members.
The lawsuit accuses OpenAI and Microsoft of engaging in unlawful and harmful conduct related to the development and operation of their AI products, including ChatGPT-3.5, ChatGPT-4.0, Dall-E, and Vall-E. The complaint claims that these products have used stolen personal information from millions of internet users, including children, without their knowledge or consent.
According to the complaint, OpenAI’s models were trained by scraping data from the internet. While web scraping is not an uncommon technique, the lack of consent and the difficulty in extracting personal information have raised legal concerns.
The complaint also states that through their AI products, defendants collect, store, track, share, and disclose private information of millions of users. This includes data collected through integrations with platforms like Slack and Microsoft Teams.
Generative AI has faced several legal challenges, including a lawsuit questioning the legality of GitHub Copilot, an AI programmer. The issue is whether Copilot’s training, which involves code from public GitHub repositories, infringes on developers’ rights or the licenses under which the repositories were made public. The case continues to proceed through the legal system.
OpenAI is not only facing legal action related to training its AI models but also defamation claims. In one instance, ChatGPT provided an inaccurate legal summary, leading to a lawsuit from a radio host who was falsely accused of financial crimes.
Regulators are also scrutinizing OpenAI’s practices. The Italian data protection regulator ordered the company to offer a right to be forgotten option to address concerns about GDPR violations.
Generative AI faces two key challenges. The first involves sourcing training data without infringing privacy or violating restrictions. OpenAI’s approach, scraping available data from webpages and tools, has raised questions about the company’s rights in gathering and using this data.
The second challenge is the trustworthiness of AI output, particularly when hallucinations occur due to insufficient data. These hallucinations were at the center of the defamation lawsuit against OpenAI.
As lawmakers strive to catch up with rapidly advancing AI technology, concerns about privacy and the use of personal data remain widespread. Some organizations, including the US House of Representatives and financial institutions, have restricted the use of generative AI due to fears of sensitive information being incorporated into AI models.
OpenAI recently updated its data usage and retention policies to allow customers to opt out of data sharing. However, these changes do not apply to data submitted before March 1, 2023, and do not cover OpenAI’s non-API consumer services like ChatGPT.
The legal challenges surrounding generative AI highlight the need for clearer regulations and guidelines to address privacy concerns and ensure responsible use of these powerful technologies.