UK launches agency to counter AI-powered disinformation from Russia

The British government has moved to counter what it calls a "poisonous tide" of foreign disinformation, announcing the creation of a new agency tasked with detecting and disrupting information warfare from hostile states. The National Centre for Information Defence (NCID), unveiled by Prime Minister Andy Burnham on Wednesday, will bring together intelligence agencies, law enforcement and social media platforms to tackle AI-enabled campaigns that officials say are increasingly targeting UK interests.

The decision reflects growing concern over the role of artificial intelligence in amplifying disinformation, particularly from Russia. While the NCID’s remit extends to all foreign threats, officials have made clear that Moscow remains the primary adversary, citing recent attempts to interfere in UK domestic politics and undermine public trust in institutions. "We cannot allow our democracy to be eroded by lies and manipulation," Burnham said in a statement, adding that the centre would "stem the flow of poison" by identifying and countering false narratives before they gain traction.

The NCID’s structure mirrors similar initiatives in the US and EU, where governments have struggled to balance counter-disinformation efforts with free speech concerns. Unlike previous UK attempts to regulate online content—such as the controversial Online Safety Act—the new agency will focus on foreign actors rather than domestic misinformation. Its mandate includes monitoring social media platforms for coordinated inauthentic behaviour, analysing deepfake content, and working with tech companies to remove or label state-backed propaganda.

Why AI changes the game

The timing of the NCID’s launch underscores the urgency of the threat. AI tools have dramatically lowered the barrier to creating convincing disinformation, allowing state actors to generate and disseminate false content at scale. A recent report by the UK’s Joint Intelligence Committee warned that deepfake audio and video could be used to manipulate elections, sow division, or even incite violence. The report highlighted a 2025 incident in which a fabricated video of a senior British minister allegedly making inflammatory remarks about immigration went viral, sparking protests in several cities.

Russia’s use of AI in information warfare has been well-documented. The country’s Internet Research Agency, a troll farm linked to the Kremlin, has reportedly shifted from manual disinformation campaigns to AI-generated content, including fake news articles and synthetic social media accounts. In 2026, UK intelligence agencies attributed a series of coordinated attacks on the country’s energy sector to Russian operatives using AI to amplify false claims about grid failures. The NCID will aim to disrupt such operations by identifying patterns in disinformation campaigns and sharing intelligence with allies.

A delicate balance

The NCID’s creation has already drawn criticism from civil liberties groups, who warn that the agency could overreach in its efforts to combat disinformation. "The line between countering foreign interference and censoring legitimate speech is perilously thin," said a spokesperson for Index on Censorship, a UK-based free speech organisation. The government has sought to allay these concerns, emphasising that the NCID will not have the power to remove content from domestic sources or penalise individuals for sharing misinformation.

The agency’s success may hinge on its ability to collaborate with social media platforms, which have faced mounting pressure to curb disinformation. Meta, X (formerly Twitter), and TikTok have all introduced policies to label or remove state-backed content, but enforcement has been inconsistent. The NCID is expected to work closely with these companies to develop real-time detection tools, though details of the partnerships remain scarce.

What’s next

The NCID will be headquartered in London, with satellite offices in Manchester and Edinburgh, and is expected to be fully operational by early 2027. Its first major test could come during the UK’s next general election, widely anticipated for late 2027 or early 2028. Intelligence officials have privately expressed concern that foreign actors may attempt to influence the vote, particularly on issues like immigration and defence policy.

For now, the government is framing the NCID as a defensive measure rather than an offensive tool. "This is about protecting our democracy, not policing it," Burnham said. But as AI continues to evolve, the agency’s role—and the broader debate over how to counter disinformation without stifling free speech—is likely to grow only more complex.