Instagram Requires AI Accounts to Self-Report

Instagram begins requiring AI accounts to self-identify and acknowledges it is not seeking to ban this content. Adam Mosseri describes ongoing moderation challenges, admits limitations in ranking these posts, and predicts an increase in hostility towards AI.

Instagram mandates identification for AI accounts and acts against fakes

Adam Mosseri states that Instagram is starting to require AI accounts to self-report, in order to help users know if what they are viewing is real. The platform had already launched a label for AI accounts earlier this year and tightened its policies against AI-generated profiles in August. Earlier this month, Mosseri clarified that Instagram is actively working to determine the right response to fake content generated by AI. He adds that it is not practical to simply ban AI content, while deceptive productions, such as celebrity deepfakes, remain problematic. This clarification came during an appearance on the Colin and Samir podcast, where he also discussed the launch of wide and long-format videos on the platform.

Ranking AI content, a recognized weak point

Adam Mosseri asserts that Instagram does not differentiate in its ranking between creators using AI and others. However, he notes a significant heterogeneity in quality among these productions, with some content being very good and some mediocre. He concedes that AI content is likely not ranked as well as non-AI content. And while he has previously used the term "slop" to refer to low-quality productions, he now believes that this term is used too freely.

Multiple concerns and growing hostility anticipated

Adam Mosseri acknowledges that a significant portion of the public does not appreciate AI, far beyond just young people and creatives. He observes increasing anxiety, frustration, and anger, particularly in creative industries, and anticipates that this negativity will intensify in the coming years, complicating the situation for many. He reports concerns regarding content ownership and model training, security issues, and the risk of job displacement, whether for content creators, engineers, or mathematicians. According to him, these discussions often reveal a broader fear of being left behind.