Meta has discontinued a new artificial intelligence feature that let users generate images by tagging public Instagram accounts, following several days of criticism over its opt-out design. The reversal matters because it reflects growing tension between AI platforms and creators over how people's likenesses are used without explicit consent, with major talent agencies and performers' unions publicly objecting. Meta said the tool "missed the mark" and was no longer available, acknowledging the feedback it had received.

The feature, powered by Meta's Muse Image model, launched on Tuesday and allowed users to manipulate an image of any person by @-mentioning their public Instagram account. Backlash centred on the "opt-out" policy, which required adult users with public accounts to actively disable the feature to be excluded. Talent agency CAA, which represents stars such as Tom Hanks and Meryl Streep, raised concerns directly with Meta, while the union SAG-AFTRA urged members to opt out; CAA later commended Meta's "swift decision" to remove it. The episode echoes OpenAI's Sora 2 video model, which had a similar opt-out feature that was criticised and eventually shut down earlier this year.