Meta has launched Meta AI – its AI assistant — in India on its major platforms, Facebook, WhatsApp, Instagram, and Messenger. While it allows users to generate text, ask questions and create a variety of tasks, closer examination of Meta’s privacy policy reveals privacy and data security concerns as Meta uses publicly available data of users and non-users, licenses online information, and content shared by users within its services to develop and train its AI model. This includes utilizing users’ public posts, images, comments, and stories from Facebook and Instagram.
The rollout of Meta AI carries privacy risks due to the enormous scale of processing user data, as is common practice for training large language models (LLMs). Importantly, as India awaits implementation of the Digital Personal Data Protection Act of 2023, it is pertinent for regulatory authorities to scrutinize automated data processing taking place en masse through AI systems. Indian users must be afforded the right to opt out of processing and such opt-out features must be simple and clear to undertake.
SFLC.in is engaged in building dialogue amongst all stakeholders to ensure the responsible development and deployment of AI systems in India.
Date: 11.07.24
Read the full statement here:
