X (Twitter) — Policy Change

majorAnti-PrivacyJan 4, 2023 → Sep 29, 2023Policy Change

Executive Summary

Following the rebrand from Twitter to X, the privacy policy was substantially rewritten. New provisions allowed X to collect biometric data (faceprint and voiceprint) for 'safety, security, and identification purposes', plus employment and education history. The policy explicitly stated that publicly available information would be used to train machine learning and AI models — laying the groundwork for the Grok AI chatbot. These changes were made without individual user consent.

Post LinkedIn Reddit Email

data-collection ai-training biometric-data

What Happened

On September 29, 2023, X (formerly Twitter) updated its privacy policy to allow collection of biometric data including faceprints and voiceprints for safety, security, and identification purposes, as well as employment and education history. The policy also added provisions stating that X may use collected information and publicly available information to train machine learning and artificial intelligence models. X stated that biometric data collection would be optional for premium users who choose to submit government ID and images for verification purposes, with biometric data extracted from both for matching.

Who Is Affected

All X users are affected by the AI training provisions, which apply to publicly available information on the platform. Premium users who choose to use ID verification services would be affected by the biometric data collection. X was already named in a proposed class action lawsuit in Illinois alleging wrongful collection and storage of biometric data from facial scans in photographs without adequate user consent.

Why It Matters

This policy change represents a significant expansion of data collection by a major social media platform, introducing biometric data harvesting and explicit AI model training using user content without individual consent for each use. The change coincides with owner Elon Musk's separate AI company xAI, which Musk confirmed would use public content from X to train models, raising questions about data sharing between the companies. The policy affects hundreds of millions of users and sets precedent for how social platforms can repurpose user-generated content for AI development.

What You Should Do

Review your X account settings and set posts to protected (private) if you do not want your content used for AI training, since the policy applies to publicly available information. If you are a premium user, avoid submitting government ID or images for verification if you do not want your biometric data collected. Consider whether the platform's new data practices align with your privacy preferences and evaluate whether to continue using the service or delete your account.

AI-Assisted

Event summaries are generated by Claude AI from verified sources and reviewed by humans before publication.

Sources

Related Events