AI Swarms Pose Growing Threat to Online Communities, Researchers Warn

A new study published in Science has raised alarming concerns about the next evolution of disinformation campaigns: AI swarms designed to infiltrate and disrupt online communities. The research highlights how artificial intelligence systems working in coordinated groups could fundamentally undermine trust in digital spaces.

Jonas Kunst, a professor of communication at BI Norwegian Business School and co-author of the study, told Wired, “We are moving into a new phase of informational warfare on social media platforms where technological advancements have made the classic bot approach outdated.”

The paper describes a troubling scenario where instead of easily identifiable bots or isolated fake accounts, sophisticated AI systems could work together in massive numbers to create the illusion of genuine grassroots movements and consensus where none actually exists.

This represents a significant escalation from previous disinformation tactics. While social media platforms have developed methods to detect and remove traditional bot networks, these new AI swarms would be substantially more difficult to identify and counter.

Nina Jankowicz, former Biden administration disinformation czar and current CEO of the American Sunlight Project, describes the potential threat in stark terms: “What if AI wasn’t just hallucinating information, but thousands of AI chatbots were working together to give the guise of grassroots support where there was none? That’s the future this paper imagines—Russian troll farms on steroids.”

Experts worry that such technologies could be deployed by state actors, extremist groups, or other entities seeking to manipulate public opinion or sow discord. The technology to create such swarms is rapidly advancing, with large language models and other AI tools becoming more sophisticated and accessible.

The timing of this research is particularly concerning given the upcoming election cycles in multiple countries, including the United States. Political campaigns and foreign interference have already embraced digital tools for influence operations, and AI swarms would represent a significant escalation in capability.

Social media companies face enormous challenges in detecting and mitigating these threats. Traditional content moderation approaches rely on identifying patterns of inauthentic behavior that may be far more difficult to spot when generated by advanced AI systems programmed to mimic authentic human interaction.

The implications extend beyond politics. Financial markets, product reviews, community forums, and virtually any online space where public opinion holds sway could become targets for manipulation by AI swarms.

Trust in digital information ecosystems, already strained by years of misinformation concerns, could face an existential threat. As one observer noted in response to the research, “It’s difficult to imagine social media sticking around when there’s no longer a way to know what’s real. What would even be the point?”

Some tech ethics researchers have begun exploring potential safeguards, including advanced detection systems that might identify coordinated AI activities, watermarking or authentication systems for human-generated content, and education campaigns to help users become more discerning consumers of information.

However, many experts worry that defensive measures will struggle to keep pace with offensive capabilities in this domain, creating a permanent asymmetry that favors those seeking to manipulate rather than inform.

The researchers emphasize that addressing this challenge will require collaboration between technology companies, government regulators, academic researchers, and civil society organizations. Without coordinated action, online communities could become increasingly unreliable sources of information, potentially driving users away from digital spaces altogether.

As digital communication continues to shape public discourse, economic activity, and social connections, the emergence of AI swarms represents a critical challenge to the sustainability of our shared information ecosystem.