NEW YORK: Once the stuff of science fiction, deep fakes are more widespread than most people realize and "democratized," according to an Advertising Week panel.
Experts on "Rage against the Machine: What Deepfakes & Algorithmic Bias Means for Your Content Marketing" tried coming to grips with the potential for harm posed by deep fakes, ranging from corporate crises to social and political discord.
Marc Minardo, lead of the digital communications group at Sard Verbinnen & Co., said deep fakes are so pernicious because of the nature of video and social media, meaning they spread quickly and are "extremely engaging."
Human intelligence is needed to understand the context around misinformation, such as what kinds of news sites are peddling it, for what purposes, who owns the sites and so on. Crises also take place in a compressed timeline, which does not lend itself to artificial intelligence, Minardo said.
"AI is developed to learn a certain amount of information over a longer period of time," Minardo added. "Technology isn’t going to have the benefit of learning."
Society is looking for a technology to weed out digital fakery, noted Gordon Crovitz, co-CEO of NewsGuard, a company that rates news sites based on trustworthiness. "There’s no way that AI, in the current state of things, can handle those issues," he contended.
To counter misinformation, an organization should deliver its messaging across all digital platforms and target key audiences in hours, not weeks, Mirando added.
If a brand’s reputation has been damaged, this targeted messaging approach would allow an organization to stifle misinformation while continuing its daily messaging, he added.
Panel members also said that media organizations suffer financially as a result of "fake news" sites siphoning programmatic advertising dollars to well-trafficked stories, sometimes as much as hundreds of thousands of dollars.