- Researchers warn AI agents could flood social media and spread false narratives
- AI swarms can mimic humans, coordinate, and sustain narratives across platforms
- Generative AI tools create credible falsehoods at scale, threatening democracy
Researchers have warned about a dystopian future where artificial intelligence (AI) agents could invade social media platforms in vast numbers and spread false narratives, harass users and undermine democracy. These 'AI swarms' will form part of a new front in information warfare, capable of mimicking human behaviour, according to a new policy forum paper published in the journal Science.
Unlike old-school bots, these AI-powered agents will be able to coordinate in real-time, adapt to feedback and sustain narratives across thousands of accounts on different platforms.
"Generative tools can expand propaganda output without sacrificing credibility and inexpensively create falsehoods that are rated as more human-like than those written by humans," the researchers stated.
Techniques used to refine AI reasoning, such as chain-of-thought prompting, could be used to generate more convincing false narratives.
"By adaptively mimicking human social dynamics, they threaten democracy. Because the resulting harms stem from design, commercial incentives, and governance."
Though the researchers did not provide a timeline for the AI invasion, they noted that the swarms would be difficult to detect, which makes the extent of their deployment unknown. The researchers said the best option to avoid such a scenario was prioritising intervention at multiple leverage points, focusing on "pragmatic mechanisms over voluntary compliance".
Watch: Man Dressed As Batman Slams US Officials About ICE Presence At Super Bowl
Social Media For AI Agents
The warning comes in the backdrop of a new social media platform called Moltbook, where only AI agents share and discuss topics, with humans allowed to observe but not post anything. This platform has been created by Matt Schlicht with the help of his AI assistant.
Schlict stated that he did not write a single line of code for the platform. He launched the site as a curiosity-filled experiment and handed it over to his AI assistant to operate. Schlicht told NBC News that he wondered what would have happened if his bot had been the founder and taken control of the entire site. He said he gave most of the responsibility to his bot, Claude Clodberg.
