Story perspectives

Critics Slam Anthropic's Controversial Claude AI Data Policy

8/30/2025

Summary

Anthropic's Privacy Policy

Anthropic's new policy mandates Claude AI users to opt out by September 28, 2025, to prevent data use for AI training, extending retention to five years for non-opt-outs.
Critics claim the policy uses "dark patterns," potentially leading to unintentional consent for data sharing, affecting all consumer tiers but exempting enterprise accounts.
The change raises privacy concerns amid reports of Claude's exploitation by cybercriminals, frustrating users over potential data misuse.

Summary

Meta's AI Privacy Concerns

Meta's "camera roll sharing suggestions" AI feature analyzes photo libraries, raising privacy concerns due to lack of user consent. It's currently available only in the US and Canada.
In response to backlash over teen safety, Meta revised its AI chatbot rules to protect minors from harmful interactions.
A lawsuit against Otter.ai claims it misuses meeting transcriptions for AI training without notifying non-account guests, potentially violating privacy laws.

History

Historical Context

The General Data Protection Regulation (GDPR), implemented by the European Union in 2018, established stringent data privacy standards requiring explicit consent for data usage. This regulation significantly impacts tech companies like Meta, which faced scrutiny for its data practices. As AI technology evolves, ongoing privacy concerns may lead to further regulatory actions, reflecting the balance between innovation and consumer protection.

1 of 3

Anthropic's Privacy Policy

Anthropic's new policy mandates Claude AI users to opt out by September 28, 2025, to prevent data use for AI training, extending retention to five years for non-opt-outs.
Critics claim the policy uses "dark patterns," potentially leading to unintentional consent for data sharing, affecting all consumer tiers but exempting enterprise accounts.
The change raises privacy concerns amid reports of Claude's exploitation by cybercriminals, frustrating users over potential data misuse.

1 / 3