The only Hong Kong-Based AXA post-doctoral fellowship awardee tackles environmental misinformation and democracy in the AI Era

People will always prioritise their beliefs, even if you show them that something is not aligned with facts and reality.

By comparing China, the United States and Europe, his study aims to show how information and ideology interact under pressure - and how different institutional settings respond when trust in institutions is strained during environmental crises.

The BRI forms the China-focused strand of the study, where Dr Karyotakis examines how geopolitical narratives and state-led communication shape public perceptions of the environmental impact of large-scale infrastructure projects. In the United States, he turns to water-pollution controversies and explores how misinformation intersects with political polarisation, regulatory distrust and long-standing inequalities in affected communities. The third case centres on European wildfires, particularly in Greece, where information and misinformation circulate rapidly in an atmosphere of uncertainty and institutional strain.

Together, the three cases allow him to probe broader global patterns: how misinformation exploits ideological divides, how it undermines democratic accountability during environmental emergencies and how different governance structures can either soften or sharpen these effects.

To make sense of these dynamics, Dr Karyotakis uses Ideological Discourse Analysis, a method that looks beyond whether a statement is "true or false" and instead uncovers the assumptions embedded in media narratives. "The approach pays close attention to who is blamed or defended in coverage of an environmental crisis," he says. "It analyses how responsibility is framed - as a failure of government, markets, foreign actors or citizens - whose voices are amplified or sidelined, and which emotions, such as fear, anger or distrust, are mobilised."

By examining online news discourse across countries, Dr Karyotakis traces how stories about risk, blame and governance move through distinct media ecosystems and how scientific evidence is reframed to fit pre-existing beliefs:

AI-driven misinformation, from synthetic images and deepfake videos to automated content, is increasingly influencing how people interpret and respond to environmental disasters. Misleading or fabricated content can confuse the public, weaken trust in experts and institutions, and delay urgent responses, especially when it is spreads through highly charged information environments.

"As misinformation becomes a tool for political manipulation, it erodes the informed deliberation that effective public decision-making depend on," Dr Karyotakis says. "This complicates evidence-based policymaking and makes coordinated climate action much harder, both nationally and globally."

Yet he also sees potential for AI to be part of the solution. "Used responsibly, AI tools can help detect patterns of misunderstanding, highlight where dialogue breaks down and elevate credible information over sensational falsehoods," he says, adding that AI can support communication strategies that resonate across ideological lines and create spaces where communities engage with environmental issues in ways that build, rather than erode, trust.

The AXA Fellowship is a funding scheme aimed at supporting young promising researchers on a priority topic aligned with AXA and the Society. Each selected researcher receives a grant of €140,000 to advance pioneering work in understanding, measuring, and mitigating the effects of misinformation.