Meta has it detected and disrupted six fresh covert influence operations from countries including China, Iran and Israel that used AI-generated content to spread disinformation.
The tech giant, which is also researching a covert Russian influence operation known as Doppelganger, shared its research on fresh online campaigns in its quarterly adversarial threats report for the first quarter of 2024, released Wednesday. He is one of many companies dealing with the the problem of disinformation AND deep fakes with the advent of generative artificial intelligence.
Many influencer campaigns were removed early on before they could build an audience of real users, Meta said, adding that it had not seen any strategies that would prevent it from shutting down the network. The company said it “observed” photos and images generated by artificial intelligence, video message readers and text. However, Meta said it has not yet observed a trend towards using photorealistic AI-generated content from politicians.
“We’re not seeing gene AI being used in a particularly sophisticated way at this point,” David Agranovich, Meta’s director of policy for threat disruption, said Tuesday during the press conference– reports Bloomberg. “But we know that these networks are inherently hostile. They intend to evolve their tactics as technology changes.
Meta did not immediately respond to a request for comment.
In its latest report, Meta said it has discovered that threat actors continue to utilize generative adversarial networks, or GANs, to create profile photos for imitation accounts, but the company can detect inauthentic networks behind these accounts.
In China, Meta said it found a network sharing photos of posters from a fictitious pro-Sikh activist movement that appeared to have been generated by artificial intelligence. Overall, Meta said it had removed 37 Facebook accounts, 13 Pages, five groups and nine Instagram accounts originating from China. The network, she said, was aimed at the Sikh community around the world, including Australia, India and the UK.
Meta said it had found a coordinated inauthentic behavior network (CIB) based in Israel that was posting comments, possibly generated by artificial intelligence, about the war between Israel and Hamas and Middle East politics on sites belonging to media organizations and public figures . Meta said the comments were mostly written in English and included “praise for Israel’s military actions” and “criticism of anti-Semitism on campus, the United Nations Relief and Works Agency (UNRWA), and Muslims claiming that ‘radical Islam’ constitutes a threat to liberal values in Canada.”
Meta removed 510 Facebook accounts, 11 Pages, one group, and 32 Instagram accounts originating from Israel that primarily targeted audiences in the US and Canada. Meta also found Iranian-based Facebook and Instagram accounts targeting Israel and containing mostly posts in Hebrew, criticizing Hamas and supporting Israel’s ultra-conservative policies.
In addition to these networks, Meta removed 50 Facebook accounts and 98 pages belonging to Bangladesh-based networks; 104 Facebook accounts, 39 pages and seven Instagram accounts based in Croatia; and 1,326 Facebook accounts, 80 pages, one group and one Instagram account from a network of unknown origin targeting audiences in Moldova and Madagascar.
