Solving the Authenticity Crisis that Social Media Created

Every morning, millions of people wake up and scroll through feeds filled with videos, images, and stories that shape their understanding of the world. A devastating earthquake in Turkey. A political rally in Iowa. A celebrity scandal in Hollywood. But here’s the unsettling reality: In our current digital ecosystem, there’s almost no way to verify whether what you’re seeing actually happened, where it happened, or if it happened at all.

We’re living through what Brandon Epstein, a digital forensic examiner and subject matter expert at Magnet Forensics, calls “the digital black hole” — a space where truth dies the moment content touches a social media platform. And the implications are staggering.

How social media confuses metadata

The problem starts with something most people have never heard of: metadata. Every photo and video you take contains invisible fingerprints. These are technical data about when, where, and how it was captured, much like a digital birth certificate that proves a file’s authenticity and origin.

But social media platforms obfuscate this evidence. “Social media sites are not actually stripping away metadata, rather creating their own images and videos based upon the submissions of their users,” Epstein explains. “It is an additive process that creates media, but does not create the same metadata values or file structure of the original.”

Think about that for a moment. When you upload a photo to Instagram, Facebook, or TikTok, the platform doesn’t preserve your original file. Instead, it creates a completely new image that appears to be what you uploaded, says Epstein.

This is an issue that has not been extensively discussed in the media. Naturally, no social media platform admits to creating a new image, but their legal terms do point to clues. For example, Facebook creates "multiple versions" of your image for different devices and compresses images to minimize storage and loading times. Instagram uses compression algorithms and may resize images to fit display requirements.

No matter the result is the same: “The posted media may appear the same as what is submitted, but it is absolutely a new image or video,” says Epstein.

This approach is baked into how these social media platforms operate. They process millions of files daily, optimizing for speed and storage efficiency. Sometimes, there are incompatibilities when sharing authenticity data. In some cases, the media may not even make the metadata available.

In most cases, forensic authenticity isn’t even a consideration. As Epstein puts it: “They are driven by the need to ingest, manage, and distribute millions of files daily — and they do so in the most efficient way they can, at the expense of forensic analysis.”

The consequences ripple outward. Every breaking news story, every viral moment, every piece of evidence that might inform public discourse gets stripped of the very data that could prove its authenticity. We’re essentially digitally blind.

The content credentials con

Then you have the issue of content credentials, which are digital signatures intended to solve everything. Major tech companies and news organizations hailed them as the silver bullet for disinformation. Adobe, the BBC, and others invested heavily in this technology, promising a future where every authentic piece of media would carry cryptographic proof of its origin.

But content credentials have turned out to be more wishful thinking than a practical solution. “Content credentials are great when the person capturing the media has the conscious intent to include those credentials,” Epstein notes. The problem? “There is no cell phone that embeds content credentials in an image/video by default, and most social media sites do not support content credentials when they create new media for posting.”

Even when platforms like LinkedIn do support content credentials, the system is weak. “It takes three mouse clicks to remove those credentials," Epstein says. For a technology designed to combat sophisticated disinformation campaigns, that’s not exactly reassuring.

The cruel irony? The most widespread use of content credentials right now comes from AI image generators, the very tools that create synthetic media in the first place. It’s like handing out certificates of authenticity to counterfeiters while ignoring legitimate creators.

Experiencing the theater of transparency

Platforms have responded to growing pressure with what can only be described as security theater. YouTube’s “generative AI” labels, for instance, rely entirely on users voluntarily checking a box to admit their content is AI-generated. There’s no independent verification, no automated detection, and no consequences for lying.

“The user has to affirmatively check a box that claims that a file is generated with AI, but there is no manner (or independent check) for any content to be flagged as ‘real,’” Epstein explains. “Even if I were to create a video and not mark it as AI, it would still be published.”

This self-policing approach would be comical if the stakes weren’t so high. Imagine if food safety worked the same way — restaurants could label their meals as “probably not poisonous” with no external verification required.

Meanwhile, the tech industry remains obsessed with deepfake detectors, which are AI tools designed to spot synthetic media. Study after study has shown these detectors are fundamentally flawed, easily fooled by sophisticated fakes while flagging legitimate content as suspicious. Yet billions continue to pour into this technological dead end.

“I do not think AI to combat AI disinformation is the answer. It is setting ourselves up for failure,” Epstein warns. It’s an arms race where the offense consistently outpaces the defense. The casualties? Truth and public trust.

The newsroom failure

Perhaps most damaging is how this broken system has infected journalism itself. Newsrooms, desperate for speed in the digital age, routinely publish unverified content from social media. They lack the forensic expertise to authenticate media properly, yet they continue to treat platforms as legitimate sources.

When asked whether journalists should refuse unverified content entirely, Epstein draws a telling parallel: “As a law enforcement officer in the U.S., if an anonymous source gave me a picture of someone who they say is selling drugs, I cannot arrest or detain that person based on that anonymous information alone.” The same skepticism that guides the criminal justice system should also apply to journalism.

Yet newsrooms continue to amplify potentially false information. During breaking news events, when emotions run high and verification time is short, this becomes especially dangerous. 

Fighting back against disinformation

The situation isn’t hopeless, but it requires a fundamental shift in how we approach digital verification. The solution isn’t more AI or better detection algorithms, but establishing a proper chain of custody for digital media from the moment of capture.

Tools like Magnet Verify represent a different approach. Instead of trying to spot fakes after the fact, they authenticate original media at the source, then use content credentials to track its journey. “It is a combination of independent authentication and content credentials,” Epstein explains.

But technology alone won’t save us. The real battle is cultural and institutional in nature. Social media platforms need pressure from newsrooms, advocacy groups, and governments to prioritize authenticity over engagement. Journalists need forensic training and tools. Users need education about how easily they can be deceived.

Most importantly, we need to accept that in an age of synthetic media, unverified content should be treated as suspect by default. The burden of proof must shift from “prove this is fake” to “prove this is real.”

The digital black hole doesn’t have to be permanent. But escaping it will require acknowledging that our current system is fundamentally broken — and having the courage to rebuild it from the ground up. The alternative is a future where truth becomes indistinguishable from fiction.

As Epstein puts it: “Every social media site wants to talk about how important combatting disinformation is, but none have taken any substantive steps to actually combat disinformation.” The question now is whether we'll force that change before it’s too late.

Image credit: iStockphoto/LightFieldStudios