Refute

Refute is the proprietary AI platform that helps governments and commercial organisations outpace their adversaries. Our responsive intelligence detects emerging disinformation campaigns before they fully manifest, and delivers context-aware response strategies so organisations can act with speed and precision. By fusing data across open-source and social frontiers and tracing narratives as they evolve, Refute transforms threat intelligence into a unified platform that aligns internal teams on decisive action. Built for those who can't afford to wait, our technology ensures you are on the front foot when it counts.

Founded in London with a Europe-wide team, Refute was built to close the gap between how fast disinformation threats move and how fast organisations can respond. Our engineers, technologists and disinformation experts develop the tools that help you detect, understand and outpace adversarial campaigns before they cause harm. Co-founders Tom Garnett and Vlad Galu have over forty years of combined experience creating world-leading data analysis systems to tackle terrorist attacks, cyber attacks, identity theft and money laundering. They have built and operated cybersecurity and identity platforms at world scale, making the internet safer for billions of people every day, and are passionate about solving the global threat of disinformation.

Refute has raised $10m to date across Pre-Seed and Seed rounds, including investors Amadeus, Playfair, Episode 1, NSSIF and Osney. As part of Refute's latest round the British government is a direct investor through the National Security Strategic Investment Fund (NSSIF), following the work Refute has carried out supporting the UK government's fight against disinformation campaigns.

Refute automatically detects disinformation behaviours that could threaten your company before they take off, including the most complex threats from foreign states using sophisticated techniques. We monitor organisations across the channels where disinformation spreads - news, social media, forums and blogs - in 100+ languages and across multimodal audio, video and text, fusing narratives to trace the flow of a campaign. Proprietary AI models detect campaigns including coordinated social media bots, and automate response suggestions: takedown requests, counter narratives, inoculation and debunking. This enables organisations to respond at the start of campaigns, before they have caused reputational and financial damage.