Prime Minister Andy Burnham stood before the UN General Assembly and delivered a pledge that set the tone for Britain’s modern security doctrine: hostile states, including Russia, which use AI-enabled information attacks and disinformation to threaten further warfare, will be met with a determined national and international response. In a speech that placed leadership at the very centre of global transformation, Mr Burnham argued that Britain wants to trade with the world and work together to solve shared problems. He was blunt, however, about the dangers of a fragmented digital age. “We need to work harder to fight the forces of division and instability which threaten,” he said. AI, in his assessment, has become a weapon of “soft division”, quietly eroding trust within societies, and that challenge has to be addressed if power is to be rebalanced in the world. The technology that can enhance lives, create prosperity and expand human capability is simultaneously capable of generating turbulence. The task, therefore, is not to reject AI but to govern it, contain its misuse and ensure national security remains robust as real warfare continues to unfold across Europe and beyond. Burnham’s address was intended to reassure allies and warn adversaries that Britain is no longer treating online manipulation as a niche concern. It is, instead, a defining feature of modern statecraft, one that requires leadership at the highest levels of government and a new kind of readiness.
At the heart of the new approach is an acknowledgment that Britain has already experienced what officials describe as an industrial-scale assault on its information environment. Foreign adversaries have waged a relentless campaign of deception, exploiting open societies and trusted brand names to inject confusion into public debate. In response, the UK government will assemble intelligence agencies, law enforcement bodies and social media companies to create an entirely new National Centre for Information Defence. The Centre has a clear remit: to detect, attribute and disrupt foreign information attacks, while helping communities build resilience to disinformation. This dual mission is crucial. Detection and attribution are necessary, but without disruption the institution risks simply cataloguing the threat. Community resilience, meanwhile, acknowledges that ordinary citizens are the final line of defence. Mr Burnham singled out Russia by name, describing the use of bots, fake websites, falsified newspaper articles and forged branding from British organisations as examples of the tactics deployed. He warned that AI will significantly increase the scale of the threat, making disinformation cheaper, faster and harder to trace. The government’s language was deliberately stark, moving beyond the idea that social media companies can police themselves. By bringing platforms into the same framework as intelligence and law enforcement, the Centre aims to act at speed and with coordinated powers.
Andrew Bud, Founder and CEO of iProov, welcomed the initiative but sounded a note of caution. “The new National Centre for Information Defence will have an important role to play, but its success will depend on how quickly it can move from identifying threats to actually disrupting them,” he said. Deepfakes, he argued, have already become too sophisticated and too scalable for the government to rely on individuals or platforms simply spotting fake content after the fact. Instead, the Centre should be looking at the infrastructure behind these attacks, including how we verify that the people, accounts and communications we interact with online are genuine. Bud also drew a direct line to the UK’s emerging cyber resilience framework. The Cyber Security and Resilience Bill recognises the need to protect the authenticity and integrity of data as part of securing essential services, and that same principle must extend to the authenticity of the people and digital interactions those services rely on. That means bringing identity verification and deepfake detection into the centre of the UK’s defensive strategy, alongside intelligence sharing and public education. “If AI is industrialising deception, our response needs to be able to tell genuine from synthetic,” he said. These remarks underscore a broader technical challenge: the difference between fighting individual falsehoods and ensuring the underlying infrastructure of online trust remains intact. The Centre, in this view, must be operational and technical as much as informational, with the authority to intervene before lies go viral.
Elizabeth Anderson, CEO of the Digital Poverty Alliance, welcomed the Prime Minister’s recognition that inequality can leave people more vulnerable to foreign influence, but she warned against framing this as a problem of individual susceptibility. People who lack reliable connectivity, the right devices, the skills to navigate online information safely, or the confidence to know what they can trust are being asked to participate in an increasingly complex digital world without the tools to do so safely. Anderson argued that as the Government develops new agencies and legislation to strengthen the UK’s resilience, digital inclusion needs to be built into that infrastructure from the outset. That means investing not only in connectivity and devices, but also in digital skills, media literacy and trusted support, alongside clear governance to ensure these protections reach the people most at risk of being left behind. A resilient digital society, she said, cannot just be one that protects its institutions from disinformation; it must also equip every citizen to participate safely within it. If new systems are built without addressing digital inequalities, there is a risk of creating resilience for some while leaving others exposed. This is a crucial caveat to the security-first language. The most sophisticated National Centre for Information Defence will not answer the problem if millions of people are unable to distinguish fact from manipulation because they lack the basic digital capabilities to evaluate online sources. Inclusion and security, in other words, must be treated as inseparable.
The international dimension was equally prominent. The UK is establishing a new AI defence partnership with the United States that will focus on protecting critical national infrastructure, signalling that the struggle against AI-enabled disinformation is now a central component of the transatlantic alliance. Such cooperation recognises that nation-states are not the only actors in this arena; criminal networks and hostile proxies can also exploit generative tools. Mr Burnham’s appearance at the UN General Assembly was therefore part of a wider effort to build a coalition of like-minded countries willing to share intelligence, set standards and coordinate responses. The involvement of His Majesty King Charles added another layer of moral authority to the debate. The King, who has long spoken about technology and society, weighed in on the enormous task before humanity, urging that AI “remains firmly in the service of humanity.” His intervention reflects a concern that goes beyond military security and into the very purpose of technological progress. The same model capable of producing convincing fake news can help treat disease, improve education or manage energy systems. The challenge is to preserve the benefits while constraining the harms. The US-UK partnership, alongside the new National Centre, is an attempt to do both, but it will need to be adaptive, transparent and accountable to avoid the very concentration of power it seeks to challenge.
In the end, Burnham’s speech and the policy announcements surrounding it amount to an argument that democratic societies must rethink what resilience means. Traditional defence rests on borders, militaries and intelligence services; information defence rests on trust, verification and social cohesion. The new Centre for Information Defence could set a global precedent if it moves quickly, integrates the private sector and treats citizens as participants rather than passive targets. But the warnings from Bud and Anderson show how easily the enterprise could fall short. Slowness to act would allow sophisticated deception to run unchecked; a purely institutional approach would leave the most vulnerable people behind. The phrase “AI remains firmly in the service of humanity” is therefore more than a royal aspiration—it is a practical test for every new agency, law and partnership. As real warfare continues to cause immense suffering, the battlefield has expanded into social media feeds, encrypted messaging and synthetic video. The UK’s promise to quash hostile information attacks will be judged by the speed and scale of its response. Success will require not just technical innovation, but a commitment to digital equality and human dignity. If AI is to be kept in service of humanity, then the defence of the truth, and of every person’s ability to recognise it, must be at the centre of the new security architecture.
