A chilling reality has emerged: AI bots are manipulating social media to sway elections. On December 14, 2025, a tragic terrorist attack in Sydney, Australia, became a breeding ground for AI-generated misinformation. As the nation reeled, social media platforms were inundated with false narratives, from manipulated videos to deepfake photos, all designed to deceive and confuse.
This is not an isolated incident. AI has become a powerful tool for spreading online misinformation globally, from Australia to Venezuela, Gaza, and Ukraine. Shockingly, nearly half of the content we encounter online is now AI-generated and disseminated. But here's where it gets controversial: these AI bots are not just spreading lies; they're creating entire fake online profiles, complete with realistic-looking social media activity, to legitimize their false narratives.
The goal? To manipulate public opinion, often for political or financial gain. But how successful are these bot networks, and how easy is it to create them? More importantly, can we, as a society, combat this digital deception through improved cyber literacy?
To find out, we conducted the groundbreaking Capture the Narrative wargame, the world's first social media simulation of its kind. We challenged students to build AI bots to influence a fictional election, employing tactics mirroring real-world social media manipulation.
Generative AI, like ChatGPT, can swiftly produce realistic text and images, making it a potent tool for creating highly persuasive fake content. Once unleashed, these AI-driven bots create a false sense of consensus by trending hashtags and viewpoints, impacting our perceptions, beliefs, and even mental health. And as bots become increasingly sophisticated, we lose faith in what we see online, creating a 'liar's dividend' where even genuine content is doubted.
Our wargame revealed startling results. Teams armed with consumer-grade AI flooded the platform with content, fracturing public debate and potentially altering election outcomes. Over a four-week campaign, competitor bots generated over 60% of the content, totaling more than 7 million posts. These bots vied for attention, unconstrained by truth, creating a manufactured reality to sway votes and drive engagement.
But the most concerning revelation was the ease with which this misinformation campaign was created. As one finalist noted, creating misinformation is scarily easy, easier than truth. This underscores the urgent need for digital literacy to empower people to identify and resist online deception.
Our simulation highlights the vulnerability of online platforms to AI manipulation. It's a wake-up call for all of us to question what we see online and to demand better tools to combat this growing threat. And this is the part most people miss: while AI can be a force for good, its misuse can have devastating consequences for democracy and public trust. So, are we prepared to face this challenge? The debate is open, and your voice matters.
