AI Products Detected
Zoom AIApple’s TrueDepth technology
AI Training Opt-Out
No opt-out available
Legal Bases for AI
consentlegitimate interestcontract performancelegal obligationvital interest
AI Usage & Disclosure
AI Disclosure

The policy explicitly mentions 'Zoom AI', 'artificial intelligence', 'machine learning', and 'intelligent features'. It also describes 'advanced tools' for automated scanning and processing of sensitive biometric data for 'enhanced features'.

Intelligent Features: Zoom provides intelligent features and products to its customers, their licensed end users and guests, such as Zoom AI or other tools to recommend chat, email, or other content, including content from third-party integrations. These tools may use artificial intelligence, machine learning, or other technology to process Customer Content and context solely to provide the intelligent features.

Zoom uses advanced tools to automatically scan certain types of content such as virtual backgrounds, profile images, incoming emails to Zoom’s native email service from someone who is not a Zoom Email user, and files uploaded or exchanged through chat, for the purpose detecting and preventing violations of our terms or policies and illegal or other harmful activity.

Enhanced Features: If you elect to enable certain features and you provide the requisite consent, Zoom may process data about your unique physical characteristics, including your facial geometry and /or voiceprint, for the purpose of offering the features.

Disclosed
Recommendation

AI tools to recommend chat, email, or other content, including content from third-party integrations.

such as Zoom AI or other tools to recommend chat, email, or other content, including content from third-party integrations.

100%
Content Processing

AI, machine learning, or other technology to process Customer Content and context to provide intelligent features.

These tools may use artificial intelligence, machine learning, or other technology to process Customer Content and context solely to provide the intelligent features.

100%
Moderation

Advanced tools to automatically scan content (virtual backgrounds, profile images, emails, files) for detecting and preventing violations of terms/policies and illegal/harmful activity.

Zoom uses advanced tools to automatically scan certain types of content such as virtual backgrounds, profile images, incoming emails to Zoom’s native email service from someone who is not a Zoom Email user, and files uploaded or exchanged through chat, for the purpose detecting and preventing violations of our terms or policies and illegal or other harmful activity.

100%
Personalization

To customize Zoom products and services and provide recommendations for users.

to customize Zoom products and services and recommendations for accounts of their users.

90%
Other

Processing of unique physical characteristics (facial geometry, voiceprint) for enhanced features, and on-device processing of face/hand movements for video effects (filters, studio effects, avatars, gestures).

If you elect to enable certain features and you provide the requisite consent, Zoom may process data about your unique physical characteristics, including your facial geometry and /or voiceprint, for the purpose of offering the features.

100%
Transcription

Generating transcriptions from audio/video content.

Content generated in meetings, webinars, messages or through other Zoom products and services... which may include audio, video, in-meeting messages, in-meeting and out-of-meeting whiteboards, chat messaging content, transcriptions, transcript edits and recommendations

90%
User Impact

AI is used for recommendations, content processing, and moderation, which can influence user experience and content. The policy also details the processing of sensitive biometric data (facial geometry, voiceprint) for 'enhanced features' with user consent. However, Zoom explicitly states it does not engage in profiling for inferring characteristics or automated decisions with legal or similarly significant effects.

Medium
Third-Party AI Vendors
Apple

The policy explicitly mentions the integration of 'Apple’s TrueDepth technology' for video effects. It also refers to 'third-party artificial intelligence models' in the context of training, indicating awareness and potential use of such models, even if not for training with customer content.

AI Training Practices
Training on Personal Data

The policy contains a clear and explicit statement that Zoom does not use 'Customer Content' (which includes various forms of personal data like audio, video, chat) to train its or its third-party AI models.

Zoom does not use any of your audio, video, chat, screen sharing, attachments or other communications-like Customer Content (such as poll results, whiteboard and reactions) to train Zoom’s or its third-party artificial intelligence models.

NO
Training on User Interactions

User interactions such as chat, screen sharing, poll results, whiteboard, and reactions are explicitly listed as 'Customer Content' and the policy states this content is not used for training AI models.

Zoom does not use any of your audio, video, chat, screen sharing, attachments or other communications-like Customer Content (such as poll results, whiteboard and reactions) to train Zoom’s or its third-party artificial intelligence models.

NO
Training on Public Content

The policy explicitly states that 'Customer Content' is not used for AI training. However, it does not specify whether other forms of public content (e.g., content from the broader internet not generated by users on Zoom) are used for AI training.

UNCLEAR
AI Data Sharing

The policy explicitly states that Zoom does not use Customer Content to train 'its or its third-party' AI models, which implies that this data is not shared for AI training purposes with third parties either.

Zoom does not use any of your audio, video, chat, screen sharing, attachments or other communications-like Customer Content (such as poll results, whiteboard and reactions) to train Zoom’s or its third-party artificial intelligence models.

NO
Risk Assessment

Despite a comprehensive privacy policy and explicit commitment not to use meeting content for AI training, the extremely low "Deterministic Score" of 0/70 indicates severe underlying privacy implementation failures or a complete lack of privacy by design. This suggests a high risk to user data, compounded by the collection of meeting content and sharing with marketing partners.

Recommended Actions

Minimize the use of Zoom, especially for sensitive conversations, until the "Deterministic Score" is clarified and significantly improved.

If use is unavoidable, ensure all privacy settings are maximally restrictive and opt out of all non-essential data sharing, particularly for marketing.

Consider alternative communication platforms with stronger, verifiable privacy track records for sensitive communications.