AI shakes statecraft: Prediction markets test diplomats, spies

As world leaders gather in New Delhi this week for the AI Summit to debate how governments should deploy artificial intelligence for public good, a less-noticed but more unsettling transformation is under way. While governments discuss how to regulate the transformative technological revolution, AI-driven prediction markets are beginning to constrain state behaviour — including on decisions about war and peace. If Trump 2.0 ushered in diplomacy breaking live as social media posts, in the age of AI, statecraft will not rest solely in the chancelleries. It will unfold in digital marketplaces where algorithms continuously calculate the odds of history even as it is made. When traders on the Polymarket began assigning probabilities to ceasefires in Ukraine, election outcomes in New York city, or interest rate cuts by major central banks, many dismissed it as digital-age gambling. But prediction markets are not just online betting platforms. They represent a new way of forecasting political and economic events — and, increasingly, a fresh force influencing them. Prediction markets are digital platforms where participants buy and sell contracts linked to future events. If a contract pays 100 rupees but is trading at 70, the market is effectively saying there is a 70 per cent likelihood of that event occurring. The basic idea is that when people risk money on their judgments, market prices tend to reveal the best informed estimate of the future. Polymarket, a blockchain-based global platform, allows users to trade on questions ranging from election results to the likelihood of war between countries. It uses cryptocurrency and operates largely outside traditional regulatory systems. For decades, geopolitical forecasting was done behind closed doors by foreign offices and intelligence agencies. They had the monopoly of classified and privileged information. Think tanks, risk analysts, and media commentariat speculated on these from outside based on their reading of open source material. Today, prediction markets are challenging state monopoly and claims of the expert class. AI models can scan vast amounts of information in real time — legislative drafts, leaders’ speeches and their career records. They can continuously monitor major decision makers, track satellite imagery, shipping data, financial flows and social media chatter. Where national bureaucracies are cumbersome, and the expert class is slow to see new patterns, AI-driven systems respond in seconds. As a result, we may be getting closer to a real-time geopolitical barometer. If markets begin assigning rising probabilities, say, to an India-Pakistan crisis turning into a war, prospects for a revolt in Iran, leadership instability in Russia, or policy shifts in China, those probabilities do not remain abstract numbers. Investors adjust portfolios and banks shift reserves. Governments feel pressure to clarify positions. Markets do not merely predict events — they can influence them. The idea that perceptions shape reality is not new. Financial markets have long demonstrated that expectations can trigger self-fulfilling outcomes. But AI amplifies this dynamic. Growing expectations of conflict may encourage military mobilisation and great power intervention. In that sense, AI-enhanced prediction markets become participants in geopolitics, not just observers. Beyond public trading platforms, a new generation of strategic technology firms is entering the field. US-based companies such as Anadyr Horizon market AI-powered geopolitical forecasting tools that simulate how political leaders might behave under stress by building their “digital twins”. Using large datasets and behavioural modelling, these systems run thousands of scenarios to estimate the likelihood of crisis escalation or diplomatic breakthroughs. Whether one hundred per cent accurate or not, they promise governments and corporations early-warning signals once available only through intelligence channels. They are also available to a wider array of actors for a price. The power of predictive markets poses major challenges to traditional statecraft. For centuries, diplomacy has been about managing uncertainty–taking big decisions in the fog of limited information. The statesmen and women relied on intelligence assessments, diplomatic advice and personal political judgement. AI-driven prediction markets introduce a new element: radical openness that thins out the fog to a large extent. Probabilities of conflict, regime change or policy shifts are continuously priced in open digital platforms. Political leaders accustomed to shaping narratives must now contend with numerical forecasts updated every second. There are clear advantages, like better situational awareness of speedier decisions. There are also dangers — AI has lowered the cost of producing synthetic information. It also increases the ease of manipulation that can generate the aura of inevitability around certain outcomes. In the AI era, information integrity itself becomes a major battleground. For emerging powers like India, the challenge is particularly acute. Governments have long operated under the assumption that sensitive information can be protected and narratives carefully managed. But in an age where probabilities of internal conflict or external crises are traded globally, secrecy becomes harder to sustain. Sound political judgement and strategies rooted in structural reality acquire a greater premium. > Abhishek Singh at Idea Exchange: ‘Whether it’s Nvidia, Anthropic, OpenAI or Google, companies are looking at India to hire AI engineers​’ Abhishek Singh at Idea Exchange: ‘Whether it’s Nvidia, Anthropic, OpenAI or Google, companies are looking at India to hire AI engineers​’ Financial markets already discipline governments. Bond markets influence fiscal policy; currency markets constrain central banks. The emerging question is whether geopolitical prediction markets will similarly restrain — or destabilise — foreign policies of major powers. Either way, they are well positioned to reshape how the world anticipates major international developments, including war and peace. AI will not replace foreign offices or intelligence agencies. But strategic forecasting is no longer the exclusive domain of diplomats and spies; it is becoming a competitive field where traders and tech firms play a growing role. For Delhi, the only option is to start building sophisticated AI-based forecasting systems of its own. C Raja Mohan is a contributing editor on international affairs for The Indian Express. He is also associated with the Motwani-Jadeja Institute of American Studies, Jindal Global University and the Council on Strategic and Defense Studies, Delhi