Deepfake “Amazon employees” are sowing confusion on Twitter
In fact, there are already some high-profile photos that have been used in harmful misinformation campaigns. In December 2019, Facebook identified and removed a network More than 900 pages, groups and accounts, Including those with in-depth profile photos, linked to the far-right issue of The Epoch Times, are known to be doing misinformation tactics. In October 2020, a fake “intelligence” document distributed in President Trump’s circles, which became the basis of numerous conspiracy theories about Hunter Biden, was also written by a fake security analyst. with a deep profile picture.
Toler says deep-seated faces have become a trend to work as open source researchers in suspicious online activities, especially since the launch of ThisPersonDoesNotExist.com, a website that offers a new face created by AI with each refresh. “There’s always a list of mind checks, every time you find anything,” he says. “The first question is,‘ Is this person real or not? “That is, we didn’t really have the question five years ago.”
What is the big threat? At the moment, Toler says the use of fake faces has not had much of an impact on his work. It’s still pretty easy to identify when a profile picture is deeply fake, just like when a photo is taken when a photo is a stock image. The most difficult scenario is the image of a real person taken from a private social media account that is not indexed in search engines.
The growing awareness of the existence of Deepfakes has led people to look more closely at the media they see, as Toler says, as evidenced by how quickly people caught up with the falsity of Amazon accounts.
Sam Gregory, director of the Human Rights Nonprofit Witness program, said this should not lead us to a false sense of security. Deepfakes are constantly “improving,” he says. “I think people will have too much confidence that it will always be possible to detect it.”
Being hyper-conscious of Deepfakes can also lead to people stop believing in real media, which can have equally serious consequences, such as weakening documentation of human rights violations.
What should we do? Gregory encourages social media users to avoid determining whether or not an image is deepfake. Often, that’s just “a small part of the puzzle,” he says. .
These research tactics are much more robust to advances in deepfake technology. This advice is also true in the case of Amazon. Checking the accounts emails and tweet details, Toler determined that they were ultimately fake. Not by looking at profile pictures.