In a series of posts on Threads, Instagram head Adam Mosseri emphasized the growing difficulty in trusting online images, citing the increasing prevalence of AI-generated content that can easily mimic reality. Mosseri warned users to be cautious about what they see online, stressing the importance of verifying sources and encouraging social platforms to play a role in helping users navigate this challenge.
“Our responsibility as internet platforms is to label AI-generated content as accurately as possible,” Mosseri wrote. However, he acknowledged that no system is perfect, and some AI-created content will inevitably go unlabeled. To address this gap, Mosseri suggested that platforms must also provide users with additional context about the individuals or accounts sharing such content. This transparency, he argued, would empower users to make informed judgments about how trustworthy the content might be.
Mosseri’s advice highlights the parallels between AI-generated images and the potential pitfalls of relying on AI chatbots, which are known to confidently present false information. He urged users to assess whether claims or visuals originate from reputable sources before believing or sharing them. This mindset could act as a safeguard against the risks posed by deceptive AI content.
Currently, Meta’s platforms, including Instagram, lack robust tools for providing the type of context Mosseri described. While Meta has alluded to upcoming updates to its content policies, details remain sparse. Mosseri’s posts hint at the possibility of new measures to help users better evaluate content, though these changes have yet to materialize.
The approach Mosseri outlined aligns with the idea of user-driven moderation seen on other platforms. For example, X (formerly Twitter) employs a feature called Community Notes, where users can collaboratively add context to tweets, and YouTube offers tools for flagging misleading content. Similarly, Bluesky has implemented custom moderation filters that allow users to tailor their experience by filtering out undesirable content. While Mosseri didn’t confirm whether Meta plans to adopt similar features, the company has been known to take inspiration from innovations on platforms like Bluesky.
As the line between real and AI-generated content becomes increasingly blurred, Mosseri’s remarks underscore the need for proactive measures by both platforms and users. By focusing on transparency and offering tools to verify the authenticity of content, social media companies can help mitigate the spread of misinformation in an era where AI-generated content is becoming more sophisticated and pervasive.