Deepfakes are manipulated videos or audio, often by changing a face, the words said, or the language spoken. Everything up until this point had been semi-automatic post-processing (mainly in cinematography). Face-swapping technology, which used to take weeks to implement, is now fully automated and takes minutes using artificial intelligence. A deepfake is more than simply an image or video; it realistically depicts an event that never happened.
Deepfakes have the potential to enable marketers in reaching out to customers with highly tailored and targeted communications. Deepfakes make it easy for influencers and celebrities to expand their reach by participating in brand promotions without showing up for a photoshoot. Imagine a new type of contract in which an influencer offers a sample of her voice as audio material and a few video shots to a brand. A brand potentially can turn that content into thousands of hyper-targeted advertisements using deepfake technology.
Zalando, a fashion retailer, executed a Facebook campaign featuring model Cara Delevingne that included 290,000 localized advertising created with deepfake technology targeted at 12 countries. The ad, produced by Infinitizer, a micro-targeting specialty firm, received 180 million impressions on social media.
Chinese technology company Tencent explained how deepfakes in the fashion industry can display apparel on a wider range of models with varying skin tones, heights, and weights. When customers try on products, they are more likely to purchase more, pay a higher price, and recommend them to their friends. The Reface AI app customers were able to visually try on Gucci items, resulting in one million face swaps in a single day.
Technology has made it more difficult to distinguish between legitimate and fabricated media. Critics warned of the risks of political deepfakes, and pornography. While the ethical implications of deepfakes have yet to be fully explored, they will continue to exist.