The UK Government has begun work on a National Centre for Information Defence, a dedicated institution tasked with detecting, attributing and disrupting hostile-state information attacks, as the Government braces for artificial intelligence to multiply the scale and sophistication of such operations. Prime Minister Andy Burnham announced the plans in a speech to the UN General Assembly in New York, where he accused Russian agencies of waging an “industrial-scale assault” on the information environments of the UK and its allies. The Centre, as reported by the BBC, will bring together intelligence agencies, law enforcement bodies and social media platforms, and will also support communities in building resilience against disinformation. Burnham said Russian operatives have deployed bots, fake websites, fabricated newspaper articles and forged branding borrowed from British organisations to spread falsehoods and undermine trust. He revealed that the branding of 28 British institutions, including universities and the BBC, had been counterfeited, and claimed the Kremlin spends around £1.3 billion ($1.7 billion) annually on information manipulation. “On top of that we have seen cyber-attacks on our companies and institutions,” Burnham told the assembly. “And we know that AI will multiply the threat.” The Prime Minister’s language was notably direct, identifying Russia by name and describing the operation as an assault rather than interference, a framing that signals a harder official line and a recognition that information operations are not a peripheral concern but a central feature of modern geopolitical conflict. By creating a dedicated centre, the UK is seeking to institutionalise a response that has previously been fragmented across intelligence, policing and regulation, and to send a message that hostile actors will face attribution and consequences for attempts to distort British public debate.
The Centre’s remit, according to Burnham, is to “detect, attribute, and disrupt” hostile-state information attacks and to defend UK democracy at a moment when the technology used to deceive the public is becoming cheaper, faster and more difficult to trace. Burnham has tasked UK security chiefs with beginning the work, signalling that the initiative is intended to operate at the highest levels of the state and to draw on classified intelligence as well as open-source analysis. The Prime Minister argued that the Government must strengthen public resilience to information warfare alongside the development of the new Centre, insisting that “security starts in every home.” That phrase captures a broader strategic bet: that citizens cannot simply be passive consumers of official messaging, but must be equipped to recognise manipulation, question dubious content and understand how foreign actors try to exploit social divisions. In practice, this is likely to mean a combination of public awareness campaigns, media literacy programmes, support for local journalism and closer cooperation with technology companies that control the platforms where disinformation spreads most effectively. It also suggests a shift in how the UK conceptualises national security, moving beyond physical threats and cyber-attacks to include the psychological and cognitive dimension of conflict. Burnham’s announcement at the UN was therefore as much about signalling to allies and adversaries as it was about domestic policy, indicating that the UK regards information defence as a permanent feature of statecraft rather than a temporary response to any single election or crisis. The timing is critical, because the next few years are expected to bring a wave of AI-enabled tools capable of generating convincing text, images, audio and video at scale, making it far harder for ordinary internet users to tell what is real and what has been manufactured by a foreign intelligence service.
The challenge of deepfakes was central to the reaction to Burnham’s speech, with technology leaders warning that the Centre will fail if it concentrates only on identifying individual pieces of fake content after they have already spread. Andrew Bud, Founder and CEO of iProov, a company specialising in face verification and deepfake detection, said the effectiveness of the Centre would depend on its ability to move beyond reactive spotting of attacks. “Deepfakes are becoming too sophisticated and too scalable for the government to rely on individuals or platforms simply spotting fake content after the fact,” he stated. Bud argued that the Centre should examine the infrastructure supporting information attacks, including how organisations establish whether people, accounts and online communications are genuine. He also called for identity verification and deepfake detection to form part of the UK’s defensive strategy alongside intelligence sharing and public education. This expands the challenge well beyond content moderation into the realm of digital authenticity, where the core question is not only “is this video real?” but “is this person real?” and “can we prove that this communication came from who it claims to have come from?” In a world where synthetic media can make anyone appear to say anything, trust in identity becomes a foundational security issue. If a hostile state can create a convincing video of a minister making a false announcement, or impersonate a chief executive to manipulate financial markets, the damage may be done before fact-checkers have even begun their work. Bud’s intervention therefore points towards a layered defence: one that combines human judgement, platform rules, authentication technology and government oversight, rather than assuming that citizens will become perfect detectors of AI-generated falsehoods on their own.
Alongside the technological challenge, there is a social dimension that risks being overlooked in the rush to build new surveillance and detection capabilities. Elizabeth Anderson, CEO of the Digital Poverty Alliance, cautioned against treating vulnerability to foreign influence primarily as a matter of individual susceptibility. She argued that people cannot be expected to defend themselves against sophisticated information warfare if they do not have the basic tools to participate in the digital world safely. “If people lack reliable connectivity, the right devices, the skills to navigate online information safely, or the confidence to know what they can trust, they are being asked to participate in a complex digital world without the tools to do so safely,” she said. Anderson called for digital inclusion to form part of the UK’s resilience planning, including investment in connectivity, devices, digital skills, media literacy and trusted support. Her comments highlight a potential blind spot in official strategy: the people most vulnerable to disinformation are often those who are already marginalised online, whether because of cost, lack of technical confidence, language barriers or limited access to high-quality information sources. A national centre focused on intelligence-led disruption may be essential for catching hostile operations at source, but it cannot solve the underlying inequality that makes some communities more exposed to manipulation. Anderson added that protecting institutions from disinformation needed to be accompanied by measures that enable citizens to participate safely online, suggesting that resilience should be understood not merely as a security objective but as a social one. Without such measures, there is a risk that the fight against information warfare becomes a top-down exercise in which the state decides what is true, rather than a genuine effort to empower a population to make informed judgements.
Burnham used the same UN speech to announce that the UK and the United States are developing an AI defence partnership intended to protect critical national infrastructure. The Prime Minister described AI as both a tool that could improve security and a “multiplier of risk”, arguing that decisions over its development require government involvement. “Fundamental decisions on this technology are too important to be left to chance,” he stressed. The partnership extends the Government’s AI security focus beyond information warfare into the protection of the systems that underpin essential services, including energy grids, water supplies, transport networks, financial systems and health infrastructure. While Burnham did not provide further operational details, the announcement signals that the UK is seeking a closer security alignment with Washington on AI, reflecting growing concern that both allied nations face similar threats from state-sponsored cyber operations and AI-enhanced espionage. The reference to critical national infrastructure is significant because attacks on such systems have traditionally been viewed as cyber-attacks rather than information operations, but the two are increasingly intertwined: a successful disinformation campaign can undermine public confidence in an institution, making it more vulnerable to a subsequent cyber-attack, while a destructive cyber-attack can be used to amplify a false narrative about who is responsible. The Government has also said it intends to put AI at the centre of the UK’s G20 presidency next year, building on the AI meeting convened by King Charles last week. Burnham stated that the UK would work towards a single set of global principles and standards covering AI safety, transparency and preparedness, an ambitious goal that would require cooperation between democracies and rivals alike in a field where technological change is outpacing regulation.
Nick Benson, CEO at CREST, an organisation that sets standards for cyber security professionals, welcomed the international focus but said global principles would only be meaningful if translated into practical, testable measures. “Policy goals and voluntary commitments alone aren’t enough to protect AI-enabled systems and critical infrastructure,” said Benson. He called for technical frameworks backed by independent assurance and practical testing, arguing that organisations deploying or evaluating AI should be able to demonstrate their capabilities and governance against assessable standards. His remarks speak to a broader concern among researchers and security professionals that government announcements in the field of AI often produce grand rhetoric but little operational clarity. For the National Centre for Information Defence to be effective, it will need not only legal powers and intelligence cooperation but also the technical means to verify claims about AI systems, test their vulnerabilities and ensure that private companies are not selling insecure or biased tools into critical public infrastructure. The announcements made by Burnham connect several areas of the UK’s developing AI security strategy: hostile-state information operations, cyber resilience, critical infrastructure protection and international standards. AI may change the scale and sophistication of those threats, but the National Centre for Information Defence is being developed against an information-warfare problem that the Government says is already active and already causing damage. The combination of intelligence-led detection, public resilience, digital inclusion, transatlantic cooperation and global standard-setting suggests that the UK is trying to build a whole-of-society response to a challenge that cannot be solved by any single agency or law. Whether that response is fast enough, well-funded enough and sufficiently trusted by the public will determine whether the Centre becomes a model for democratic governments around the world or another institutional reaction to a threat that evolves faster than the institutions designed to meet it.