ARTICLE AD BOX
New Delhi: With AI playing an increasingly prominent role in daily digital interactions, technology firms are encountering a fresh challenge: establishing user trust. AI tools are changing rapidly, but users and regulators are asking questions about the use of their personal data and whether they have any control over it.
The balance became apparent this week when Meta pulled a recently launched AI image-generation tool shortly after its release. The tool was part of the company’s new Muse Image model, which enabled people to create AI images based on photos from public Instagram accounts. It was soon criticized for privacy and consent issues, and Meta discontinued the feature.
Muse Image is the first image generation model created by Meta Superintelligence Labs, the company’s dedicated AI division. The model was launched earlier this week and is now accessible via the Meta AI app, Instagram Stories and WhatsApp in certain markets, and as part of the Meta AI chatbot. It could also edit images based on sketches and annotations, in addition to creating images from text prompts.
The debate focused on a functionality that enabled users to create AI images by tagging public Instagram accounts. Critics said the design was a burden on users, not requiring explicit consent, as public accounts were automatically eligible unless users opted out via privacy settings. However, privacy experts cautioned that this capability could be exploited, such as through impersonation, deepfakes and non-consensual manipulation of images.
The privacy backlash was not limited to privacy advocates. The union of actors and media professionals SAG-AFTRA has condemned the feature, arguing that AI systems using people’s likenesses should be clearly opt-in. The organisation subsequently applauded Meta’s decision to take the functionality away, saying it was the right thing to do in the face of the dangers of digital replicas.
In response to the criticism, Meta admitted that the feature was “missed the mark. The company stated that it was planning to provide a creative tool but also wanted to give users the choice over whether their public content could be referenced, but it confirmed that the feature had been discontinued.
The key is how fast the turnaround is. Before changing, large technology firms tend to defend their new products. In this instance, Meta pulled the feature from its website within days of its release, demonstrating the rapid pace at which public opinion can sway AI product decisions.
The episode is also a sign of a wider trend in the tech sector. As companies strive to develop more sophisticated generative AI products, privacy expectations are rapidly changing as well. Even if a feature meets platform policies, it is under more scrutiny if it uses publicly available user content.
AI has been a key focus of Meta’s future plans, with the company investing heavily in new AI models and user-facing tools through its Superintelligence Labs. The image generation and editing tool, Muse Image, is still there, but the option to use public Instagram accounts has been taken away.
The incident is a reminder in many ways that technical ability is no longer sufficient. With the growing adoption of AI across platforms that billions of users rely on, companies are increasingly expected to design products from the ground up with transparency, consent and user trust in mind. The rapid rollback is a sign that for Meta, these are now as critical as implementing the next AI feature.
English (US) ·