Instagram’s chief executive, Adam Mosseri, has said that the rise of artificial intelligence-generated content is leading to a future where it may be more practical to identify real media rather than fake media. In a recent post, he noted that the assumption that photographs or videos accurately capture real moments is becoming obsolete, stating, “This is clearly no longer the case and it’s going to take us years to adapt.” He emphasized that the traditional approach to labeling AI-generated content is becoming outdated, as technologies like watermarks have proven unreliable. Instead, he suggests that camera manufacturers should develop systems to verify authenticity at the moment of capture, potentially transforming how users discern genuine images from AI-generated ones. Mosseri advocates for better creative tools, labeling AI content, and verifying authentic media to help users navigate this evolving landscape. Mosseri also noted that the polished aesthetic of Instagram posts is fading, with a shift toward more raw and unfiltered images that can help creators demonstrate authenticity in an increasingly AI-dominated landscape.
Why do we care?
We’re moving toward a world where reality needs receipts.
When the head of Instagram says it might be easier to prove something is real than fake, that’s an admission that trust is gone by default. And once trust is gone, someone has to rebuild it—and whoever does gets power.
Camera-level verification sounds reasonable until you ask who controls it. Which devices qualify? Which platforms recognize them? And what happens to content that can’t be verified—not because it’s fake, but because it came from the wrong place?
The MSP and IT services mistake here would be thinking this is a social media problem. It’s not. This is coming for security footage, compliance evidence, training data, marketing assets—anything where “this really happened” matters.
And the real risk isn’t deepfakes. It’s outsourced judgment—letting platforms decide what counts as real while you’re still responsible for the consequences.
This matters now because once “provable real” becomes expected, the absence of proof will be treated as suspicion. The providers who understand that shift early won’t just help clients adopt tools—they’ll help them navigate a world where trust has become technical, brittle, and expensive.