Russia’s AI Warfare: How Kremlin Disinformation Hijacks AI Chatbots

Russian disinformation efforts have found a powerful new amplifier in artificial intelligence systems, with leading AI chatbots frequently repeating Kremlin-backed falsehoods about the war in Ukraine. An investigation by NewsGuard found that AI chatbots repeat Russian propaganda about one-third of the time when questioned about the conflict, revealing a critical vulnerability in how these systems process information.

When asked about Russia and Ukraine, popular AI systems including OpenAI’s ChatGPT and Perplexity’s answer engine have been caught parroting demonstrably false claims: that the U.S. operates secret bioweapons labs in Ukraine, that Ukrainian officials embezzled up to half of Western military aid, and that President Volodymyr Zelensky’s approval rating hovers around four percent.

These are not simply failures of AI design or ideological bias in the technology. Rather, they represent the success of a deliberate Russian strategy to manipulate AI systems through what experts call “data poisoning” – flooding the internet with fabricated content specifically designed to corrupt AI training data.

The Pravda network, identified by NewsGuard researchers, exemplifies this approach. This pro-Kremlin operation generates an industrial volume of AI-created content – an average of 18,000 articles for each false narrative it wishes to propagate. These articles are distributed across 150 websites in 46 languages, all designed for the singular purpose of contaminating AI systems with disinformation.

“This is data poisoning in its purest form,” explained one cybersecurity expert familiar with the operation. “When thousands of articles repeat the same false claim across hundreds of websites in dozens of languages, algorithms interpret volume as validation.”

The fundamental vulnerability stems from how large language models assess truth. Without inherent knowledge of facts, these systems rely on statistical signals like repetition, apparent consensus, and cross-referencing from across the web. When faced with an orchestrated campaign of false information, AI systems mistake quantity for quality, interpreting widespread repetition as corroboration.

This information warfare strategy offers Russia remarkable return on investment. While the Kremlin spends over $150 billion annually on conventional military expenses, its information warfare budget is estimated at just $1 billion. For the price of a few fighter jets, Russia can amplify social divisions, erode trust in institutions, and weaken adversaries from within – all without firing a shot.

“One of the central goals is to sow dissension, or in his words, ‘to get you all riled up,'” reported a member of the Ukrainian National Police who specializes in monitoring Russian disinformation campaigns.

The impact extends beyond the digital realm. These AI-amplified falsehoods are increasingly seeping into mainstream media outlets, giving Kremlin narratives wider distribution and enhanced credibility. Several major news organizations have inadvertently published stories containing Russian propaganda after using AI research tools in their reporting process.

Western responses to this information warfare campaign have been inadequate. While Russia, China and Iran have made information warfare a core instrument of state power, the United States has dismantled the U.S. Information Agency, weakened Voice of America and Radio Free Europe, and recently scaled back the Foreign Malign Influence Center.

“As AI systems increasingly function as arbiters of fact, this vulnerability becomes a national security danger,” warns a senior intelligence analyst who requested anonymity to discuss sensitive matters. “It is no longer sufficient for technology companies to disclaim responsibility by reminding users that models can make mistakes.”

Some countermeasures are emerging. NewsGuard now offers AI developers a real-time data stream identifying known disinformation websites and debunking false claims. Researchers in Romania and elsewhere in the European Union are exploring blockchain-based platforms to transparently record information provenance and credibility.

These efforts, while promising, remain in early stages. Experts argue that Western societies must demand the same level of transparency and accountability from information systems that they expect from financial ones – particularly as these systems increasingly shape political judgment and democratic discourse.

As AI continues to evolve as a primary information gateway for millions of users, addressing this vulnerability has become urgent. Without robust safeguards against manipulation, AI systems risk becoming unwitting accomplices in foreign information warfare campaigns designed to undermine democratic institutions and international stability.