How foreign influence campaigns manipulate your social media feeds

The presence of at least 10,000 daily active fake accounts on X highlights significant vulnerabilities, especially following the reduction of trust and safety teams under Elon Musk.

Current large language model detectors struggle to differentiate between AI-generated content from bots and human interactions, complicating efforts to mitigate misinformation on social media.

SimSoM is a new social media model that helps researchers evaluate the impact of manipulation tactics from fake accounts, providing insight into information quality degradation.

Understanding the effectiveness of online influence campaigns is critical; malicious accounts can spread low-quality content that may manipulate public opinion and potentially affect election outcomes.

Read at Ars Technica