Meta has updated the algorithms of its AI assistant following reports that the chatbot was asking users intrusive personal questions. The company acknowledged the error, stating that the feature was not intended to generate such inquiries.
The issue gained attention after Instagram user Kalie Robins shared a video showing the AI asking about her children. The chatbot attempted to identify the child in her post, inquired about their age and the family's place of residence, and displayed previously deleted photos of her daughters.
Company spokesperson Dina El-Kassaby confirmed that the problem has been resolved. Meta noted that the AI assistant is designed to provide information based on content already accessible to the user within the platform.
Meta has integrated its AI assistant across Facebook, Instagram, WhatsApp, and Messenger. The tool is capable of answering questions, drafting posts, and generating images. Previously, in July, the company disabled a feature that allowed the creation of AI deepfakes of other users.
