Instagram’s AI image generator alarms privacy experts

Summary

Meta has released Muse Image, an AI image-generation model that lets users reference public Instagram profiles by tagging them in prompts. The system extracts facial features and other visual elements from posts on those profiles to synthesize new images, without notifying the individuals whose content is used. Public accounts are included by default; only private accounts and those belonging to users under 18 are automatically excluded.

The opt-out control resides in the “sharing and reuse” section of Instagram settings and must be toggled manually. Privacy researchers at the Electronic Frontier Foundation and cybersecurity firms including Malwarebytes and Proton argue that this design places an unreasonable burden on users and that the setting is difficult to locate. They note that the toggle states are visually similar, increasing the chance that the feature remains active unintentionally. Thorin Klosowski of the EFF described the approach as inappropriate for a novel use of long-published photographs.

Meta states that Muse Image incorporates safety guardrails and that adult users with public accounts can disable reuse of their content. The company has not clarified whether images of minors appearing in adult public posts are protected from reference. The tool is currently accessible via the Meta AI app, Instagram Stories in the United States, and WhatsApp in selected countries, with further rollout planned for Facebook and video generation.

Why it matters

This article is highly relevant for privacy professionals as it highlights a major shift in how social media platforms utilize user data for AI generation. In the context of the Dutch and EU market, this default opt-in approach raises severe GDPR compliance questions and necessitates immediate policy reviews for enterprise social media usage.