However, the form does not account for a user's activity on Meta-owned properties, such as Facebook comments or Instagram photos, meaning the company could potentially use such first-party data to train its AI models. A Meta spokesperson clarified that their latest Llama 2 open-source large language model was not trained on Meta user data and that they have not yet launched any Generative AI consumer features. The spokesperson also noted that depending on where people live, they may be able to exercise their data subject rights and object to certain data being used to train Meta's AI models.
Key takeaways:
- Meta, formerly Facebook, has updated its help center to include a form that allows users to delete some personal information used in the training of generative AI models.
- The form refers to third-party information, which is data publicly available on the internet or licensed sources, and does not account for a user's activity on Meta-owned properties.
- A Meta spokesperson stated that their newest Llama 2 open-source large language model wasn't trained on Meta user data and they have not launched any Generative AI consumer features on their systems yet.
- Data privacy advocates have raised concerns about the practice of aggregating vast quantities of publicly available information to train AI models, prompting a joint statement from data protection agencies to remind tech companies of their obligations under various data protection and privacy laws.