Microsoft warns of Russian online campaigns targeting US election

Concerns about AI manipulation are raised, with Microsoft finding that simpler digital forgeries are more common than deepfakes and audio manipulations have a greater impact than video.

Lumen Technologies linked a cyberattack on U.S. internet firms to a Chinese government-backed group.

According to Microsoft, Russian online campaigns aimed at influencing the upcoming US presidential election have been detected over the past 45 days, albeit slower than previous election cycles. These Russia-linked activities involve disseminating divisive content, particularly criticising American support for Ukraine in its conflict with Russia, targeting US audiences. The tech giant’s researchers highlighted a rise in messaging concerning Ukraine through traditional media and social platforms, involving at least 70 Russia-affiliated activity sets.

The most prominent of these campaigns is reportedly linked to Russia’s Presidential Administration, while another focuses on spreading disinformation across multiple languages. This strategy involves initially posting content on video channels under the guise of a whistle-blower or citizen journalist. The disinformation is then amplified through a network of websites such as DC Weekly and Miami Chronicle. Microsoft noted that US audiences often need to know its origins before sharing and reposting this disinformation.

Microsoft also identified an increase in hacking activities by a Russian group, Star Blizzard (or Cold River), targeting Western think tanks. This hacking campaign, which has shifted its focus to US political figures and policy circles, is potentially laying the groundwork for Kremlin objectives leading up to the November elections. Despite concerns about the malicious use of AI, Microsoft emphasised that simpler digital forgeries, particularly audio manipulations, are more prevalent than deepfakes and can significantly impact audiences.

While acknowledging nation states’ use of generative AI-enabled content, Microsoft researchers highlighted that such content rarely achieves substantial reach on social media. Basic manipulations, rather than complex AI techniques, will likely be the most impactful in influencing public perception. Microsoft’s findings underscore ongoing concerns about foreign interference and disinformation campaigns targeting the US electoral process.