Skip to content
Submit incident
AII-1002-01357
ReportedReported by fewer than three independent publishers, with no official finding.Supporting sources are pending verification.

Meta AI requests access to private user photos for image analysis

Meta launched a feature requesting access to users' private photo galleries stored on their mobile devices, claiming it analyzes images to generate editing suggestions. Privacy experts raised concerns that Meta's terms of service do not clearly restrict future use of the photos for AI training, unlike comparable services offered by competitors.

28 June 2025 Β· Meta Β· Privacy and surveillance
28 June 2025
AII-1002-01357
Quick facts
Organization
Meta (subject)
Category
Privacy and surveillance
Country
United States
Date
28 June 2025
Drafted by
Raidu Incident Detection

What happened

Meta began displaying a new feature in Facebook that requests access to private photos stored in users' mobile galleries. The feature appears when users create a story and uses the Meta AI blue circle icon to request permission to upload private images to Meta's servers for analysis and automatic generation of edited versions, even if the photos have not been published on the social network.

By accepting, users authorize Meta to continuously upload multimedia files and analyze them using contextual information such as time, location, and visual elements. Users also agree to Meta AI's Terms of Service, which authorize automated analysis of photos, facial recognition, and use of metadata.

Reported impact

Privacy experts warned users to exercise caution with personal data shared through this feature. Meta stated that the edited suggestions are optional, visible only to the user unless they choose to share them, and that content will not be used to train AI models during this testing phase.

Index analysis

Meta's terms of service permit the company to 'retain and use' personal information, and while the company asserts the feature is not currently used for AI model training, the conditions do not exclude that possibility in the future. Google Photos with Gemini generates similar visual suggestions but Google explicitly states it does not use personal data to train its generative models, whereas Meta has not provided this guarantee.

Governance controls

Meta confirmed through spokesperson Maria Cubeta that the feature is in testing in the United States and Canada. The company stated the suggestions are optional and can be disabled at any time.

Classification

Organization
Meta (subject)
Category
Privacy and surveillance
Industry
Media, social platforms and marketing
Country
United States
Harm status
Potential harm
Intent
Unknown
Timing
Post-deployment
Harmed
Consumers
Failure modes
Training data misuse
AI systems
Meta AI

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident