The Chief Policy Officer at Meta Platforms mentioned to Reuters during an interview that the company has used public posts on its platforms Facebook and Instagram to train its new virtual assistant (Meta AI), which is based on artificial intelligence. However, the social media giant has excluded private posts exchanged with family and friends in its efforts to respect user privacy. Meta's Global Affairs President, Nick Clegg, stated that the company also did not use private chats from its messaging services as training data for the model and has taken steps to filter private data from the public databases used in training.
He added on the sidelines of the company's annual Connect conference this week, "We tried to exclude databases that include massive amounts of personal information," noting that "the vast majority" of the data Meta used for training was available publicly. He cited LinkedIn as an example of a website from which Meta chose not to use content for privacy reasons.
*Intellectual Property Dilemma*
Companies are evaluating how to handle proprietary or protected materials that have accumulated in the process, which AI systems might potentially reproduce, as they face lawsuits from data creators accusing them of copyright infringement. Meta stated that it built the assistant using software, services, and materials based on the LLaMA 2 large language model it released for commercial use.
The assistant will be able to generate text, audio, and images, and will have access to real-time information through a partnership with Microsoft's Bing search engine. Clegg mentioned that regarding proprietary materials, he expects "a reasonable amount of litigation regarding whether creative content is covered by the current fair use principle," which allows limited use of protected works for purposes such as commentary, research, and parody.
A Meta spokesperson pointed out that the new terms of service prevent users from generating content that violates privacy and intellectual property rights, in response to a question about whether Meta has taken steps to avoid reproducing copyrighted images.
