Meta says its new AI image tool blocks explicit content, but consent isn’t part of the design

Meta says its newly launched Muse Image tool blocks violent, sexual or defamatory imagery of real people, with content that breaks its Community Standards subject to enforcement. Underage public accounts are walled off from the feature entirely, and teenagers cannot use it to generate images of others.

What the safeguards don’t address is consent itself. Muse Image, which launched Tuesday out of Meta Superintelligence Labs, lets anyone tag a public Instagram handle and generate AI images built from that account’s published photos and reels — with every adult public account automatically opted in and no notification sent when someone remixes their likeness.

Meta AI chief Alexandr Wang promoted the launch on X as a creativity tool for custom event invites, mock-ups and personalised graphics, describing it as “personal.” The tool is genuinely capable, sitting at No. 2 on the Arena leaderboard for text-to-image generation — but critics say the core design choice, letting the feature run on strangers by default, undercuts Meta’s own safety messaging.

Users can opt out through Instagram’s settings menu, under “Sharing and reuse,” by switching off content reuse for Posts and Reels — though images generated before opting out remain live regardless.