Meta is under scrutiny again, this time over its use of user photos for its AI systems on Facebook and Instagram. The company is expanding its data collection efforts to include photos stored on users’ devices, even those never shared online.
The new cloud processing feature prompts users to grant Meta access to their entire photo galleries. Meta states that this is to provide creative suggestions. However, it raises significant privacy concerns as it allows the company to access and analyze unshared photos.
When users enable this feature, Meta can examine faces, objects, and contextual details within the photos. Meta argues this is an opt-in feature, but it’s questionable whether users fully comprehend the extent of data sharing.
Meta has previously admitted to utilizing public photos uploaded since 2007 for AI training. The ambiguity surrounding what constitutes ‘public’ and the use of data from users who were minors in 2007 raise ethical concerns.
Users can disable this feature within their Facebook settings. If disabled, Meta claims it will delete the unshared photos from its cloud servers within 30 days. The issue is that this feature is being presented as a helpful tool, but it is actually quietly expanding the boundaries of user privacy.
As AI permeates numerous areas, companies are increasingly driven to gather more user data. Meta’s new practice raises questions about user privacy and whether seemingly convenient features can jeopardize it. Users are urged to review their Facebook settings, specifically looking for options like ‘Cloud Processing’ or ‘Camera Roll Access,’ and disabling them if necessary.