<img src=”https://storage.googleapis.com/prod-zenger-upload/image/20230906/feat_607aa4fa-b646-4642-b678-fbba49514205.jpg” alt=”Meta (Facebook) sign is seen at its headquarters at Menlo Park in California, United States on August 5, 2023. Facebook’s Privacy Center contains a recently announced privacy feature that explicitly covers Generative AI Data Subject Rights. Nevertheless, it only applies to what Meta refers to as third-party data acquired from outside sources through data scraping, purchases, or license agreements. TAYFUN COSKUN/ANADOLU AGENCY/GETTY IMAGES“>
Today, when data privacy concerns are paramount, Meta Platforms, the parent company of Facebook and Instagram, has introduced a new privacy setting that prompts a critical question: Is the company’s data request form a genuine safeguard against the intrusion of artificial intelligence into users’ personal information?
What Happened: The newly unveiled privacy feature, buried deep within Facebook’s Privacy Center, specifically addresses “Generative AI Data Subject Rights.”
However, it only pertains to what Meta refers to as “third-party data” obtained from external sources through data scraping, purchases, or licensing agreements, reported Gizmodo.
To make a request, users need to provide only their name, email address, and country of residence. The company stated that it aims to protect user privacy, but the approach raises questions about whether it can effectively identify and exclude all relevant data tied to an individual.
The outcome of requests submitted through the form varies depending on local laws and regulations.
In regions with robust data protection laws, such as the U.K. and Canada, there may be stronger safeguards against data usage for AI training.
In contrast, in countries like the U.S., where federal regulations on AI and privacy are less stringent, the effectiveness of this form may be limited, the report noted.
For the unversed, while Meta’s AI models are not solely reliant on user-generated content within its platforms, the information users input directly on Facebook, Instagram, or other apps may still be fair game for AI training, regardless of their preferences expressed through the newly introduced form.
In 2021, Facebook was accused of prioritizing its financial interests over the safety of its users. At the time, speculation emerged suggesting that hackers had placed the personal data of more than 1.5 billion Facebook users up for sale online.
Produced in association with Benzinga