Contractors involved in training Meta's AI reported accessing sensitive user data, such as names, phone numbers, and personal conversations, while reviewing chatbot interactions. These workers claimed that they often handled chats containing personal information, including selfies, which raised significant privacy risks. The contractors noted that users frequently shared personal topics in their discussions with the AI, heightening concerns about data security. While similar projects have been conducted by other tech giants, the reviewers emphasized that Meta's tasks appeared to include more personal details than those from other companies.
Contractors hired to train Meta's AI reported seeing personal user data, including names, phone numbers, and intimate conversations, raising serious privacy concerns.
Four contract workers disclosed that they frequently encountered personal details within thousands of chats they reviewed, pointing to a significant data privacy risk in AI training.
Collection
[
|
...
]