Instagram's Mosseri tells users: if you don't like AI content, don't use the feed

Instagram head Adam Mosseri said during an interview on Lenny Rachitsky's podcast that "I don't think we should filter out AI content," instead arguing that the platform "should let you know if content is AI content or not."

The most controversial part of his message: if a user does not want to see AI content, the solution is in their hands - scroll past it, do not engage, or hide individual posts.

The Verge reported on these statements on July 10, 2026, the same day Meta removed the "Muse Image @mention" feature from Instagram following a massive backlash from users.

The timing reveals an internal contradiction. Meta wants to embed AI tools throughout Instagram while avoiding the responsibility of filtering the AI content those tools produce. Mosseri's stance delivers a clear message to brands and creators: Instagram won't protect your organic reach from AI slop. You're competing for feed space on equal terms with synthetic content, whether you like it or not.

What Meta removed (and why it matters)

Meta announced Muse Image, a new AI image generator built by Meta Superintelligence Labs, promoting one feature that allowed individuals to generate images by @-mentioning public Instagram accounts that they wanted to reference.

The feature worked by allowing users to reference public Instagram accounts in their prompts, pulling photos from those accounts into newly generated images. Rather than requiring users to actively agree to participate, the feature enrolled all public accounts automatically.

The reversal followed criticism from users, privacy advocates, actors, and talent representatives over its automatic opt-in approach and the risk that people's likenesses could be reused without clear permission. SAG-AFTRA urged its members and other Instagram users to opt out shortly after the feature launched.

The company issued a blog post Friday announcing that it was removing the feature, stating "Our intent was to provide a useful creative tool and to give people control over whether their public content could be referenced in this way."

The retreat exposed Meta's strategic tension. The company is investing heavily in generative AI tools, actively encouraging users to create synthetic content, while simultaneously facing resistance when those tools cross consent boundaries. Mosseri's refusal to filter AI from feeds looks less like a principled stance on user choice and more like a calculated compromise: pull back the most controversial features, but preserve the AI-saturated feed.

The "label it, don't filter it" policy

Instagram is committed to placing clear labels on AI content, indicating that it was "created or edited using AI." Instagram introduced a new "AI Creator" label for accounts that regularly produce content using artificial intelligence tools in May 2026, allowing qualifying creators to toggle the label on their profile.

Instagram has clarified that content ranking, distribution and recommendations remain unchanged. This means creators are not penalised or boosted based on the tag.

That's the mechanism brands need to understand. AI content isn't deprioritised. It isn't relegated to a separate feed. It competes for the same algorithmic distribution as human-created work, with the only distinction being a label most users will scroll past without noticing.

Mosseri believes people who love AI content "should be able to have a feed that's just AI town." The implication: Instagram's algorithmic recommendation system will serve you more of what you engage with, whether that's human or synthetic. If you don't want AI, just don't interact with it. The problem for brands is that users don't always know what they're engaging with until after the fact.

Mosseri said in the podcast episode that aired on Thursday: "In a world where there's an abundance of synthetic content, I actually think people are going to seek out creativity and authenticity and people more, not less."

That's a theory. The reality for social media managers is that they're fighting for reach in a feed where "the feeds are starting to fill up with synthetic everything," and the platform won't help them differentiate their content beyond a label.

What this means for brand reach in 2026

According to data from Socialinsider, the average Instagram post now reaches 3.5 to 7.6% of a brand's followers in 2026 (down from 10 to 15% in 2020), with a year-over-year organic reach drop of 12% from 2024 to 2025.

Three reasons combine: content saturation (too much competing content), platform monetisation limiting unpaid reach, and the algorithm's shift from showing posts to followers toward AI-driven recommendation.

AI content accelerates all three. Every AI-generated post from a creator or aggregator account adds to the saturation. Every synthetic image or video competes for the same feed slots your brand content needs. And because content ranking and distribution remain unchanged regardless of the AI label, brands get no algorithmic advantage for being human.

The signals that drive reach in 2026 have shifted. Watch time is the number-one ranking factor for Instagram content, confirmed by Adam Mosseri in January 2025. Mosseri confirmed "sends per reach" is weighted 3-5x higher than likes in the 2026 algorithm. Likes now carry minimal weight. Instagram downgraded likes from 20% weight in 2023 to just 5% in 2026.

AI tools are optimised to generate content that holds attention. Synthetic videos can be edited to maximise watch time. AI-generated images can be engineered to trigger saves and shares. If those signals are what the algorithm rewards, and AI content can produce them at scale, brands running human-led content strategies are competing at a structural disadvantage unless their work delivers genuinely superior engagement.

The broader context: Mosseri's AI stance

Mosseri has been signalling this direction for months. In a post looking forward to 2026, he wrote: "one major shift: authenticity is becoming infinitely reproducible. Everything that made creators matter - the ability to be real, to connect, to have a voice that couldn't be faked - is now accessible to anyone with the right tools."

He says there is "a lot of amazing AI content" and that the platform may need to rethink its approach to labeling such imagery by "fingerprinting real media, not just chasing fake."

That statement deserves unpacking. If Instagram moves to fingerprinting real media rather than labelling synthetic content, the default assumption shifts. Instead of AI being the exception that needs a label, human-created content becomes the exception that needs verification. That's not a hypothetical. It's the logical endpoint of Mosseri's position.

Mosseri points out that most users under 25 no longer care about posting to their public grid. Instead, they're sharing "unflattering candids" in private group chats or disappearing stories. That shift reflects a deeper behavioral change: people don't trust how things look anymore - they trust who's posting them.

For brands, that creates a strategic fork. Either invest heavily in cultivating trust and recognisable identity so your audience knows your content is human, or accept that you're competing in a feed where synthetic content is treated as equivalent and optimise accordingly.

What to do about it

Instagram won't filter AI content. Labelling doesn't affect distribution. Organic reach is already constrained. Here's what social media professionals can actually control:

Prioritise the signals that matter. Instagram's 2026 algorithm rebalanced distribution around four buyer-intent signals: DM shares, saves, watch time on Reels, and profile clicks. Posts that generate those signals in the first 20 minutes get pushed to wider audiences. Design content to trigger shares and saves, not just passive likes.

Lean into recognisable human voice. If users are shifting trust from what content looks like to who's posting it, your brand voice and identity become defensive moats. Make your content unmistakably yours. Consistency in tone, perspective, and creative choices signals authenticity in ways a label can't.

Test Reels aggressively. Reels reach roughly 2.25× more people than single-image posts, and about 55% of their views come from non-followers. If you're trying to break out of a shrinking follower-based reach, Reels are still the primary discovery mechanism.

Accept that volume alone won't save you. Around 5 to 7 posts per week is the sweet spot. Posting more than 10 times weekly shows diminishing returns, while fewer than 3 posts a week tends to hurt engagement and reach. Quality of engagement signals beats frequency.

Consider whether Instagram is still the right channel for your objectives. If your strategy depends on organic reach to your existing followers, and that reach has fallen to single digits, either adapt to recommendation-led distribution or move resources to channels where you still control access to your audience.

Mosseri's stance isn't going to change. Meta has made its bet on AI-infused feeds, and it's willing to absorb user backlash on individual features like Muse Image while holding the line on the broader policy. Social media marketers need to decide whether they're willing to compete in that environment, and if so, what they're changing to win in it.