Security researchers have uncovered a sophisticated Russian disinformation campaign known as “CopyCop,” which leverages generative AI (GenAI) to manipulate and disseminate biased content from major news organizations. This operation, identified by Recorded Future, employs large language model (LLM)-powered GenAI to replicate legitimate news content and then inject political bias to target specific audiences. Renowned media outlets such as Al-Jazeera, Fox News, the BBC, La Croix, and TV5Monde have been affected. In some instances, legitimate sites are spoofed and hosted on alternative domains, while other times, the fabricated stories are published on fictitious news sites like London Crier and GB Geopolitics.
The narratives propagated by CopyCop aim to advance Russian influence by creating division over issues like the Israel-Hamas conflict, undermining support for Ukraine, and criticizing governmental policies. Recorded Future reported stories falsely claiming that the UK government has criminalized Islam and plans to establish a ‘buffer zone’ of NATO countries around Ukraine. Another goal is to drive a wedge between the UK and US governments.
The potential impact of such GenAI-powered disinformation is particularly concerning ahead of key elections in the UK and US. Clément Briens, a threat intelligence analyst at Recorded Future, emphasized the unprecedented reach and effectiveness of AI in shaping public opinion. He urged media organizations and government entities to proactively identify and dismantle these influence networks and work with naming authorities and hosting providers to block infringing domains. The report also highlighted the importance of exposing these operations and communicating their impacts accurately to the public. Collaboration between governments, tech companies, and civil society is essential to combat the spread of online disinformation, ensuring the integrity of Western democracy and trust in mainstream media.
Reference:
