Meta Pulls Instagram AI Feature After Deepfake Backlash

Meta has switched off an Instagram feature it launched only days earlier, after the tool — which let anyone generate AI images from a public account's content just by tagging it — drew significant backlash.

What happened

The feature let users create AI-generated images based on content pulled from any public Instagram account, triggered simply by tagging that account. As originally built, it meant content from any public account could be used to generate AI images without the account owner's permission. [VERIFY]: The feature's official product name, its exact launch date, and the full mechanics of how tagging triggered image generation.

Meta turned the feature off this week following what it described as significant backlash.

Why it matters

For account holders: the core complaint was consent. A public Instagram account is, by definition, visible to anyone — but "visible" and "usable as raw material for AI-generated images of you, triggered by a stranger's tag" are different things. Meta's decision to build a no-opt-in pathway by default, rather than an opt-in one, is the part critics zeroed in on.

For Meta: this is the latest instance of a now-familiar pattern across the industry — ship an AI feature fast, absorb the backlash, walk it back. That cycle carries a cost beyond the immediate reversal: it erodes trust that the next AI feature Meta ships has actually been stress-tested for misuse before launch, not after.

For the broader AI industry: deepfake-adjacent tools keep landing in the same trap — technically impressive, legally murky, and reliant on users to police their own restraint. Any platform sitting on a large corpus of public images and video is one product decision away from this exact scenario. Regulators watching the deepfake space will likely note that even Meta, with its scale and legal resources, still shipped this without a permission gate.

What's unknown

[VERIFY]: How long the feature was live before removal, how many images were generated during that window, whether Meta plans to reintroduce an opt-in version, and whether any regulators or lawmakers have commented publicly.

The takeaway

The pattern here matters more than the specific feature. Consent defaults are a design decision, not an afterthought — and "public account" is doing a lot of load-bearing work as a justification for using someone's likeness in AI-generated content without asking first. Expect this incident to become a reference point the next time a platform proposes a similar tagging-triggered AI tool, and expect more scrutiny on whether "opt-out after backlash" is an acceptable substitute for "opt-in by design."

Source: The Verge