Fake social media accounts are increasingly likely to feature fake faces.
Meta, the parent company of Facebook, claims that computer-generated profile pictures were used in more than two-thirds of the influence operations it discovered and shut down this year.
Bad actors are adapting the artificial intelligence behind these fakes for their attempts to manipulate social media networks as it becomes more widely available and better at creating faces that look real.
Ben N-I M O, who is in charge of Meta's global threat intelligence, stated, "It looks like these threat actors are thinking, this is a better and better way to hide."
This is due to the fact that it is simple to simply go online and download a fake face instead of stealing an entire account or photo.
They probably thought, "It's a person who doesn't exist, so nobody will complain about it and people won't be able to find it in the same way," N-I M O stated.
The fakes have been used to spread propaganda from China and Russia and to harass activists on Twitter and Facebook. This year, an NPR investigation revealed that marketing con artists are also using them on LinkedIn.
The innovation behind these countenances is known as a generative ill-disposed network, or GAN. Although it has existed since 2014, it has significantly improved over the past few years. Nowadays, anyone can generate fake faces for free or for a small fee on websites.
According to a study that was published earlier this year, AI-generated faces are now so convincing that people only have a 50% chance of correctly identifying whether a face is real or fake
However, computer-generated profile pictures frequently include recognizable characteristics that people can learn to recognize, such as strangely aligned eyes, odd clothing, and strange backgrounds.
He stated, "The human eyeball is an amazing thing." When you take a gander at 200 or 300 of these profile pictures that are created by computerized reasoning, your eyeballs begin to recognize them.
That is made it simpler for specialists at Meta and different organizations to recognize them across informal communities.
The threat actors are in a paradoxical position because they believe that by employing these pictures generated by AI, they are concealing something very cleverly. "But in fact, they're actually throwing up another signal, which says, this account looks fake, and you need to look at it", he said. "But in fact, to any trained investigator who has those eyeballs skills, they're actually throwing up that signal".
He asserts that this is a significant factor in the evolution of threat actors since 2017, when Facebook began publicly eradicating networks of fake accounts attempting to covertly influence its platform. Since then, it has shut down more than 200 of these networks.
He stated, "We're seeing online businesses just trying to spread themselves across more and more social media platforms, not just the big ones but also as many of the smaller ones as they can". This includes popular petition websites in addition to new and alternative social media platforms like Getter, Truth Social, and Gab.
"Threat actors are just trying to put their content in different places." He added, "And I think it's in the hope that something won't get caught."
Meta claims that because threats rarely exist on a single platform, it collaborates with other tech companies and governments to share information about them.
But the work with a crucial partner's future is now in doubt. Under the ownership of Elon Musk, Twitter is undergoing significant change. He has drastically reduced the trust and safety workforce at the company, including non-English language teams and state-backed propaganda operations. All the key leaders in trust, security, privacy, and safety have left.
Nathaniel Gleicher, the head of security policy at Meta, stated, "Twitter is going through a transition right now, and most of the people we've dealt with there have moved on". Therefore, we must await their announcement in these threat areas
You must be logged in to post a comment.