Ratings / AI App
Critical Rating 1 / 5 · Critical — Active surveillance / serious risk.
Anthropic's Claude poses significant privacy risks due to its data collection and sharing practices. In August 2025, Anthropic shifted its consumer data policy to an opt-out model for using conversations to train AI models, extending retention for opted-in data to five years. The
Anthropic's Claude poses significant privacy risks due to its data collection and sharing practices. In August 2025, Anthropic shifted its consumer data policy to an opt-out model for using conversations to train AI models, extending retention for opted-in data to five years. The privacy policy states, "We may use your Inputs and Outputs to train and improve Anthropic AI models, unless you opt out through your account settings." Furthermore, in June 2026, the policy was updated to allow proactive sharing of conversation data with law enforcement based on a "good-faith belief" without a court order, as detailed in Section 3: "We may share personal data with government authorities, law enforcement... where... we have a good-faith belief that disclosure is reasonably necessary." This stance was underscored in March 2026 when a federal judge ruled that communications with Claude were not protected by attorney-client privilege due to its privacy terms. Additionally, in September 2026, a class action lawsuit was filed regarding the requirement for biometric scans for identity verification, which the policy confirms can include "facial geometry templates."
The full methodological and legal context for this rating. This card is what we point to when asked to substantiate the claim.