When Instagram head Adam Mosseri shared a 20-slide reflection on the platform’s future on December 31, 2025, he focused on a central idea: “authenticity is becoming infinitely reproducible.” Mosseri said AI has altered the meaning of photos and videos online, as synthetic media becomes increasingly difficult to distinguish from real moments. He argued that trust in platforms like Instagram will increasingly depend on who is sharing content rather than what the content looks like.

For most of Instagram’s history, images served as proof of real moments. Filters and editing tools existed, but they enhanced reality instead of replacing it. A photo started with something that actually happened, whether a meal, a travel shot, or a candid moment. "For most of my life I could safely assume photographs or videos were largely accurate captures of moments that happened,” he said.

According to Mosseri, that assumption no longer holds. He said the world is entering an era of “infinite synthetic content,” where AI systems can generate images and videos that look indistinguishable from real life. The result, he warned, is a slow collapse of visual trust.

“We are going to move from assuming what we see is real by default to starting with skepticism. We need to pay attention to who is sharing something and why. This will be uncomfortable because we are genetically predisposed to believing our eyes,” he said.

What Instagram used to be and what changed

In the early 2010s, Instagram was defined by the public feed. Square photos. Carefully framed meals. Edited travel shots. Beauty filters, but still required a camera and a real subject. Even when editing tools improved, content was still anchored to reality. A creator could enhance a photo, but they still needed to take one. The image began with a moment that actually happened.

However, Mosseri says that version of Instagram is gone. “Unless you are under 25, you probably think of Instagram as a feed of square photos,” he wrote. “That feed is dead. People stopped sharing personal moments on the feed years ago.” Users gradually stopped sharing personal moments on the feed and moved those interactions to Stories and private messages.

How AI tools are shaping Instagram content

The turning point came as generative AI tools moved from novelty to mainstream. When OpenAI released DALL-E in 2021, it demonstrated that images could be generated from text alone. Midjourney and similar tools followed, enabling creators to produce surreal or cinematic images without a camera or a real subject. By 2023 and 2024, platforms like Runway made it possible to edit or generate videos with the same ease as photos. 

In August last year, Google launched Nano Banana, which allowed stylized images to be created and shared.  Meta also has ‘Movie Gen’ that creates short videos from text prompts and has partnered with Midjourney to bring AI image tech to future products.

These AI-created images and videos began appearing alongside real content on Instagram. A sunset, a portrait, or a travel clip might be entirely generated, making it increasingly difficult for users to distinguish reality from AI. Mosseri says the platform was built on the assumption that captured media represented reality, but AI breaks that assumption.

Why imperfection became popular

As AI imagery improved, creators leaned into raw, imperfect visuals. Blurry photos, shaky videos, and poor lighting became signals of authenticity. In Stories and direct messages, casual, unpolished content felt more real than highly edited feed posts. Mosseri said, “In a world where everything can be perfected, imperfection becomes a signal. Rawness is not just an aesthetic preference anymore. It is proof.”

But Mosseri argues this phase will not last. “Relatively quickly, AI will create any aesthetic you like, including an imperfect one that presents as authentic,” he warned. AI systems are already learning to replicate imperfection. Grain, blur, uneven lighting, and even shaky motion are becoming reproducible. When that happens, aesthetics will stop working as proof.

From trusting images to trusting people

Mosseri predicts that judging content will shift from visuals to context. Users will need to consider who is sharing content and why, rather than whether an image looks real. “We are going to move from assuming what we see is real by default to starting with skepticism,” he wrote. Identity, consistency, and context will become key measures of trust.

What Instagram says it needs to do

According to Mosseri, Instagram will explore several strategies. He says the platform must adapt by surfacing credibility signals. That includes clearer labeling of AI-generated content, verification of authentic media, and more information about the accounts behind posts.

The platform will expand automated and manual labels that let users know if something was created or altered by AI. This helps viewers understand the origin. However, Mosseri is clear that platforms cannot solve this with labels alone. “Labeling is only part of the solution,” he wrote.

Instagram plans to continue labeling AI-generated content, but Mosseri admits platforms will get worse at detecting AI as models improve. That is why he suggests verifying real media instead. “It will be more practical to fingerprint real media than fake media,” he said, pointing to cryptographic signing by cameras at the moment content is captured.

Mosseri says ranking systems should highlight content that reflects a creator’s original voice and real contribution, not recycled or deceptive media. “We need to surface much more context about the accounts sharing content so people can make informed decisions,” Mosseri wrote.

What this means for creators

Mosseri believes that AI content will not reduce the importance of creators. He says there is “a lot of amazing AI content,” and Instagram is actively building creative tools that use AI. “Authenticity is becoming a scarce resource,” he wrote. “Driving more demand for creator content, not less.” When anyone can generate a realistic image or video, the value shifts. The question is no longer whether someone can create content. It becomes whether they can create something that only they could make and maintain trust over time.

Creators who rely on aesthetics may struggle. Those who build consistent voices, show process, and share context may stand out more as audiences grow more skeptical.

The risk for Instagram

However, Mosseri says the biggest risk Instagram faces is failing to keep up with how quickly the world is changing. AI is not coming; it is already integrated in how content is made, edited, and shared.

As Instagram heads into 2026, Mosseri believes the platform must evolve fast or risk losing relevance in a feed where nothing can be taken at face value. Authenticity, once assumed, now has to be proven.

‍