As major tech companies scramble to tackle “deepfakes” — audio and video content manipulated by artificial intelligence — technology giants like Facebook (FB) and Google (GOOG) are fighting a “losing game,” one San Diego-based company said on Tuesday.

Speaking to Yahoo Finance Truepic CEO Jeff McGregor said the rapid advancement of A.I., and the scale at which visual deception is being disseminated, poses major challenges to researchers developing tools for detection.

Instead of trying to detect what’s false, McGregor stated companies should focus on establishing what’s real.

“There are just too many fundamental issues with the way photos are captured and stored today to really be able to do any effective detection at internet scale,” McGregor told “On The Move” on Tuesday.

“You need a human, you need a forensics expert to be able to do that. That just can’t work at the scale of a social media company,” he added.

Deepfake content has proliferated at a rapid rate, as online tutorials and forums increasingly democratize A.I. tools once reserved for tech’s elite users. The total number of manipulated content nearly doubled in 2019, with roughly 15,000 posts shared online, according to a recent study by Deeptrace Labs.

Last fall, Facebook and Microsoft teamed up with university researchers to launch the Deepfake Detection Challenge, to develop tools to detect fake content.

“If Facebook, and Google and other large tech companies are releasing data sets and asking for help, candidly that means the top AI researchers in the country, if not the world have not figured out how to detect deepfakes,” McGregor said. “It’s a very telling sign.”

When you can’t believe what you see

Truepic own technology uses Controlled Capture to lock in a photo or video the instant it is created. That allows users to authenticate the pixel data of an image and certify the time, date, and location at the moment of capture.

Each image comes with a URL, so anybody capturing the content can prove that it’s real in the future.

The digital tool has already gained traction with a number of enterprise customers — particularly insurance companies like Jewelers Mutual Insurance, who use Truepic’s technology to authenticate photos included in claims.

The company has also worked closely with non-governmental organizations (NGOs) and the State Department, to train citizen journalists in the Middle East. In 2018, the technology got widespread attention after a Syrian teenage activist used Truepic to record a verified message to President Donald Trump, calling on him to “protect children like me” from Bashar Al Assad’s chemical attacks.