Meta Reports Russia’s AI Tactics for US Election Interference are Ineffective

Department of Research, Strategic Studies and International Relations 16-08-2024

A new security report from Meta reveals that Russia’s attempts to use generative AI for online misinformation campaigns have largely failed to achieve significant results.

The parent company of Facebook and Instagram reported on Thursday that AI-driven strategies have so far only yielded “marginal productivity and content-generation benefits” for malicious actors. Meta has been successful in disrupting these deceptive influence campaigns.

Amid growing concerns that generative AI could be exploited to mislead voters during elections in the United States and other countries, Meta continues its efforts to combat “coordinated inauthentic behavior” across its platforms.

Russia remains the primary source of such activities, relying on fake Facebook and Instagram accounts, said David Agranovich, Meta’s security policy director. Since Russia’s 2022 invasion of Ukraine, these efforts have focused on undermining Ukraine and its allies.

As the US election draws nearer, Meta anticipates that Russia-supported online misinformation campaigns will target political candidates who back Ukraine.

For years, Facebook has faced criticism for being a potent tool for election-related disinformation. Russian operatives previously used Facebook and other US-based social media platforms to stoke political divisions, notably in the 2016 election won by Donald Trump.

Experts are increasingly worried about a surge of misinformation from bad actors on social networks, facilitated by generative AI tools like ChatGPT or the DALL-E image generator, which can create content rapidly.

The report states that AI has been employed to produce images, videos, translations, and fake news stories. When searching for deceptive practices, Meta focuses on account behaviors rather than the content they post.

These influence campaigns often span multiple online platforms, and Meta has observed posts on X, formerly known as Twitter, being used to lend credibility to fabricated content. Meta collaborates with X and other internet companies, emphasizing the need for a united front against misinformation.

Regarding X’s cooperation, Agranovich noted, “As far as Twitter (X) is concerned, they are still undergoing a transition. Many of the contacts we previously worked with have moved on.”

X has significantly reduced its trust and safety teams and scaled back its content moderation efforts, which once helped manage misinformation, turning it into what researchers now consider a hotspot for disinformation.