Meta, in a recent security report, says Russia is employing generative artificial intelligence (AI) tools for online deception campaigns about the US elections. The tech giant claims most of the efforts were unsuccessful, as it foiled the campaigns.
According to Meta, the generative AI strategies have brought minimal productivity and gains. This is because the company could detect and clamp down on coordinated inauthentic behavior (CIB) on its platform.
Meta identifies Russia as the primary source of political influence campaigns
The Meta Adversarial Threat Report for Q2 identified that most CIB campaigns on its platforms, particularly Facebook and Instagram, originated in Russia. Since 2017, the country has been responsible for 39 covert operations, while Iran and China come second and third with 30 and 11 operations, respectively.
These Russian campaigns relied heavily on fake accounts and pages, with one core tactic being acquiring multiple accounts. In terms of content, they all appear to have been propagating the same message since Russia invaded Ukraine in 2022—justifying the invasion.
Meta has now predicted that more campaigns targeting US users will surface in the next few months as the November presidential election draws near. The campaigns will likely spread anti-Ukraine messages and attack candidates supporting Ukraine.
It said:
“Between now and the US elections in November, we expect Russia-based operations to promote supportive commentary about candidates who oppose aid to Ukraine and criticize those who advocate for aiding its defenses.”
Despite the persistent threat and the heavy use of AI in these deceptive campaigns, Meta claimed it had developed a highly effective way of detecting and preventing such malicious activity. The company focuses on the account’s behavioral pattern instead of the post’s content, enabling it to detect automated actions better.
Russia’s history of election meddling
Meta’s proactiveness in limiting malicious political influence campaigns is not surprising. The company has come under fire from regulators for failing to prevent or remove such campaigns from its platform. In May this year, the European Union opened an investigation into Meta’s failure to comply with its election integrity rules.
Meanwhile, Russia has a history of sponsoring influence campaigns against rival world powers, particularly during their elections. The country has been accused of influencing the outcome of the 2016 US elections through campaigns across every social media platform, including apps like Tumblr and PayPal.
A study analyzing millions of social media posts across all platforms, including Google, showed that the Russian government, through several operatives, spread misinformation about the US electoral process to undermine it. It identified one group, the Internet Research Agency (IRA), considered by many to be a Russian troll farm, as leading the campaign.
With the arrival of generative AI technologies, which make generating all forms of content easier, there have been concerns that their use for influence campaigns would be very high in this election. So far, several security experts have identified some instances of its use. However, even political candidates have accused each other of using AI.
