arXiv:2605.30166v1 Announce Type: cross Abstract: LLM-driven social bots can generate fluent, human-like text, reducing the discriminative advantage of content-based detection alone. However, coordinated campaigns still leave relational patterns -- interactions, behavioral similarity, shared neighborhoods, community positions, and coordinated activity -- that graph-based methods can exploit. Existing graph detectors face two challenges when exploiting such evidence. First, Euclidean GNNs distort hierarchical and scale-free social graphs; while hyperbolic geometry addresses this volume-growth m
The proliferation of advanced LLMs has significantly empowered social bots, making their detection more challenging through traditional content analysis methods.
This research offers a new computational approach to identify sophisticated AI-driven disinformation and manipulation, which is critical for maintaining robust social and political discourse.
The ability to more accurately identify LLM-driven social bots shifts the battlefield from content analysis to relational patterns and network geometry for detection.
- · Social media platforms
- · Cybersecurity firms
- · Democratic institutions
- · Research in graph neural networks
- · State-sponsored disinformation campaigns
- · Malicious actors using AI bots
- · Traditional content-based bot detection methods
Improved detection capabilities will disrupt large-scale coordinated influence operations on social media.
Adversarial AI development will accelerate to create bots that can evade these new graph-based detection methods.
The arms race between AI bot development and detection will continue, potentially leading to more advanced and covert forms of information warfare.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG