Instagram chief says fingerprinting real media is more practical than labeling AI fakes

AI content dominates Instagram feeds

Adam Mosseri, the executive leading Instagram, observed that the platform’s visual stream is increasingly filled with synthetic imagery created by artificial intelligence. He described this trend as a fundamental shift, noting that the tools enabling anyone to produce realistic images are now widely accessible.

Challenges of labeling AI‑generated media

Mosseri explained that existing approaches—such as watermarks or platform‑level labels—have proven unreliable. Watermarks can be removed or ignored, and Meta’s own labeling system does not consistently identify AI‑generated or manipulated content. He warned that as AI models improve, detection will become even harder.

Fingerprinting real media as a practical alternative

Instead of chasing every fake, Mosseri advocated for a system that “fingerprints” authentic media at the moment of capture. He suggested that camera manufacturers could embed cryptographic signatures in photos and videos, creating a chain of custody that verifies the content’s origin. This approach, he argued, would make it easier for users to trust what they see.

Implications for creators and photographers

The shift toward AI‑filled feeds has already frustrated many creators who rely on Instagram for exposure. Mosseri noted that the classic “polished” aesthetic is fading, and that raw, unflattering images may become a new hallmark of authenticity. He implied that creators who embrace this style could better demonstrate that their work is genuine.

Future outlook

While acknowledging the difficulty of the problem, Mosseri remains confident that a collaborative effort—particularly involving camera makers—could provide a viable solution. He emphasized that the platform’s responsibility may be less about policing every piece of synthetic media and more about empowering users to recognize verified real content.