Mosseri Expects AI Backlash on Instagram to Worsen Over Two Years
Adam Mosseri said AI negativity is not only getting worse but will continue to get worse, predicting more frustration and anger over the next couple of years. He made the remarks on the Colin and Samir podcast, where he discussed the launch of widescreen and long-form video for Instagram.
During the interview, Mosseri was asked about AI and the trend of college students booing tech CEOs at graduation ceremonies. He acknowledged that many people really dislike AI, and that the hostility is not confined to young people and creatives. He also described a rise in anxiety, frustration and anger about AI, particularly in creative industries.
On distribution, Mosseri said Instagram does not rank AI creators differently from other creators. He also said Instagram probably does not rank AI content as well as non-AI content. The two statements address different things: the first concerns how creators are classified, the second how their output performs in the feed.
On enforcement, Mosseri said Instagram is now starting to require AI accounts to label themselves, so people can tell whether what they are looking at is real or not. He said he does not think it is practical for Instagram to try to ban AI content from the platform. The approach is disclosure rather than prohibition, and the rollout is at an early stage, as the phrasing of the remark indicates.
Mosseri also said the term "AI slop" is used too liberally. The phrase has become shorthand for low-effort generated content, and his objection is to its application across a broader range of material than the label was coined to describe.
The broader context here is the gap between the two halves of Mosseri's remarks. A platform can decline to ban a category of content while still wanting viewers to know what that content is. Disclosure systems depend heavily on how accurately creators self-report, and on what happens when they do not. Labelling also does not resolve the ranking question he raised. A viewer who knows a post is generated may still engage with it less, and the labels do not change the ranking signals the feed uses to decide what to show.
In my view, the most useful reading of the interview is that the platform is accepting the backlash as a fixed condition and designing around it, rather than expecting it to fade. A prediction of worsening sentiment over two years is a forecast, not a finding, and the public reaction to AI will depend on factors well beyond any single platform's policy.
Still, a visible label gives viewers a concrete tool for sorting what they see, and that has value regardless of how the broader sentiment moves. Whether disclosure can ease the anger Mosseri describes is a separate question, and the next two years of feed behaviour, creator response and viewer reaction will be where it gets answered.
