Paragraph 1: The New Face of an Old Threat
Election interference is hardly a new phenomenon. For as long as nations have held elections, other nations have tried to influence their outcomes, whether through espionage, propaganda, or diplomatic pressure. But the game has changed, and it has changed in ways that make the old playbook seem almost quaint. The United Kingdom, like many democracies, now faces a world where artificial intelligence can craft disinformation at a scale, speed, and level of sophistication that human propagandists could never match. The National Cyber Security Centre has sounded the alarm: emerging technologies, particularly AI-generated content, are cheaper than ever, capable of reaching millions of voters with terrifying precision, and much harder to detect than traditional forms of interference. This is not a distant problem or a hypothetical scenario. It is happening now, in real time, across social media platforms and messaging apps, often hidden behind fake accounts and murky bot networks. The danger is not simply that a voter might see a misleading post. The danger is that the entire information ecosystem becomes polluted, leaving ordinary citizens unsure of what is real and what is a fabrication. In such an environment, trust in democracy itself begins to corrode. That is why it is so important for the UK to have a clear legal framework for understanding and responding to these new tactics. Vague policy statements and ad hoc responses are no longer enough. We need to know where the lines are, what is lawful, what is unlawful, and how the international community can hold bad actors accountable. This report is about drawing those lines, not in the abstract language of international law, but with the practical aim of protecting the integrity of British elections and the democratic process as a whole.
Paragraph 2: The Danger of AI Disinformation and the Search for Legal Clarity
The central finding of the research is both straightforward and deeply concerning: AI-generated disinformation poses a serious and immediate danger to UK elections. But here is the puzzle. There is no specific international law rule that says, in plain terms, “election interference is illegal.” International law was largely developed in an era of physical borders, diplomats, and state-to-state relations, not in an era of deepfakes, algorithmic manipulation, and viral bots. So how do we address a problem that the old rules never anticipated? The answer, according to this research, is that we are not without tools. The ancient and fundamental principle of non-intervention in international law provides a firm foundation for declaring that AI-generated disinformation is unlawful. The non-intervention rule, at its heart, is about respect for sovereignty. It says that no state has the right to interfere in the internal affairs of another state. It says that coercion is forbidden, and that states cannot use their power to force other states, or their people, into making decisions they would not otherwise make. This is not just a political principle; it is a binding rule of international law, recognized by courts and states around the world. The challenge is applying this rule to a world of tweets and videos, where the coercion is not a navy blockading a port or an army massing at a border, but a flood of fabricated information designed to manipulate the minds of voters. The law has to adapt, and the non-intervention rule offers a way forward. It may not have been written with artificial intelligence in mind, but its underlying logic captures the essence of the problem: when one state uses deceptive content to strip another nation’s citizens of their ability to make free and informed choices, that is an intervention in the most profound sense.
Paragraph 3: A Practical Test for Unlawful Disinformation
How can we tell the difference between legitimate political discourse and unlawful intervention? The research offers a clear, two-part test based on the non-intervention rule. The first question is about deception. Was the information communicated with the intention of deceiving the audience into believing that something false is actually true? This is not about stating an opinion or putting a positive spin on a policy. This is about deliberately creating content that is designed to fool people, to make them think a candidate said something they never said, or that a policy proposal contains provisions it does not contain, or that a political figure is involved in a scandal that never happened. AI makes this terrifyingly easy. A realistic video of a politician saying something outrageous can be created in minutes. A fake audio recording can be produced with just a few samples of a person’s voice. The second question is about coercion. Would a reasonable observer conclude that the communication was intended to influence the audience’s decision-making to such an extent that they would be left without a meaningful choice about what to think and therefore what to do? This is the key to understanding coercion in the digital age. Coercion does not always mean physical force. It can also mean manipulating the environment so that a free choice is impossible. If a voter is bombarded with fabricated scandals, false promises, and manipulated evidence, their decision is no longer truly theirs. They are being herded toward a conclusion by a foreign power that is pulling the strings behind a curtain of fake accounts and algorithmically amplified lies. If both answers are yes, then the AI-generated disinformation is a violation of the non-intervention rule. It is not just misleading; it is a coercive act, intended to get the target audience to do something they would not otherwise do, and to leave them with no real choice in the matter.
Paragraph 4: What Is Unlawful and What Is Not
Applying this legal framework to real-world scenarios reveals a clear hierarchy of unlawful conduct. First, state cyber influence operations that use AI-generated disinformation are, with very few exceptions, unlawful as a matter of international law. When a foreign government deliberately creates and disseminates false content designed to manipulate the electoral choices of another nation’s citizens, that government has crossed a clear line. It does not matter if the operation succeeds or fails. The mere act of attempting to coerce another state’s population through deception is an illegal intervention. Second, and even more clearly, state operations that promote racial or religious hatred are always unlawful. This is a principle that goes beyond the non-intervention rule. It touches on fundamental norms of human rights and human dignity. Incitement to hatred is not simply a form of election interference; it is an attack on the very fabric of a society, designed to sow division, encourage violence, and undermine the equal standing of citizens. International law is absolutely clear on this point, and there is no room for debate. However, there is also an important boundary that protects free speech. The non-intervention rule does not forbid states from placing true information into the public domain. If a foreign government publishes accurate facts about a candidate’s criminal record, or exposes a genuine corruption scandal, that is not intervention. It might be unwelcome, it might be politically motivated, and it might even be intended to influence an election, but it is not unlawful. There is a crucial distinction between truth and lies, and the law must preserve it. We cannot allow the fight against disinformation to become an excuse for suppressing uncomfortable truths. The line is not between foreign speech and domestic speech; it is between deceptive manipulation and honest information.
Paragraph 5: What the UK Government Should Do Next
Knowing the law is one thing; acting on it is another. This report offers practical recommendations for the UK Government, and these are not soft suggestions. They are necessary steps to protect the democratic process. First, the Government should support a major public information campaign to explain the problem and the illegality of state-sponsored disinformation campaigns. The public needs to understand that this is not just about fake news or partisan bickering. It is about foreign powers attempting to undermine their ability to vote freely. An informed public is a resilient public. If citizens understand the tactics being used against them, they are far less likely to fall for them. Second, the Government must clarify its position on international law in relation to AI-generated disinformation targeting elections. It should make an unambiguous statement that AI disinformation is illegal under international law, and it should set out the legal reasoning behind that position. This is not a matter of academic interest. It is a matter of deterrence. Foreign states are more likely to engage in these operations if they believe there are no clear costs. A clear, public, legally grounded position sends a message that the UK will hold states accountable for their actions. Third, the Government should work with its international partners to establish and codify the rules of international law in this area. The non-intervention rule is a strong foundation, but it needs to be explicitly applied to the new realities of cyber operations and AI. The UK should lead the way in building a global consensus that election interference through AI-generated disinformation is a violation of sovereignty and will not be tolerated.
Paragraph 6: The Researcher Behind the Report and an Open Invitation
The research behind these findings was led by Professor Steven Wheatley, a Professor of International Law at Lancaster University. His work focuses on two principal areas: human rights theory and democracy, and this report sits at the intersection of those two interests. The protection of elections is not just a matter of legal technicality; it is a matter of human rights, and it is fundamental to the concept of democracy itself. Professor Wheatley’s work is not meant to remain locked in an academic journal. It is a call to action for policymakers, for legal scholars, for civil society organizations, and for anyone who cares about the integrity of democratic elections. The problem of AI-generated disinformation is developing rapidly, and the law must keep pace. But this cannot happen in isolation. It requires dialogue and collaboration across disciplines and across borders. If you would like to learn more about this research, invite Professor Wheatley to speak at your event, or explore potential collaboration, please contact him directly at [email protected]. The challenge is immense, but it is not insurmountable. By bringing together legal expertise, technological awareness, and public engagement, we can build a framework that protects our elections from those who seek to undermine them. Democracy has survived war, depression, and division. It can survive the age of artificial intelligence, but only if we take the threat seriously and act with clarity and purpose. The law is one of our most powerful tools, and it is time to use it.