Overview
- Meta introduced Muse Image on July 7 as an image generator that let users reference any public Instagram account by @-mention, then removed the account-reference capability after heavy backlash.
- The pulled feature worked by default, requiring account holders to opt out rather than opt in, and it did not notify people when their public photos were used as inputs for AI images.
- Creators, SAG-AFTRA and privacy groups warned the default setting raised the risk of non-consensual likeness use, deepfakes and scams because the tool made it easy to generate realistic images of real people.
- Separately, advertisers report that Meta’s Advantage+ Creative Enhancements and the Andromeda system have auto-modified live ads without clear consent, producing distorted images and incorrect copy in campaigns.
- The two episodes underline a broader risk from Meta’s rapid push to automate ads and expand AI features, and they increase the chance of regulatory scrutiny, legal challenges and demands for clearer consent and controls.