Cracking the Code: How AI Detects Social Media Impersonation

Cracking the Code: How AI Detects Social Media Impersonation

Author: Osnat Mokryn

The rise of online communication and the increasing anonymity of internet interactions have led to growing concerns around authorship attribution and impersonation detection. Social platforms like Twitter, Facebook, and Reddit have further opened the door to deceptive behaviors—ranging from fake identities and manipulated reviews to coordinated misinformation campaigns. Identifying who is behind a post—or whether the same individual operates multiple accounts—has become increasingly challenging. A new AI-driven approach offers a breakthrough by analyzing the words people tend to use and the ones they consistently avoid. This technique constructs a unique linguistic “signature” for each user, enabling accurate authorship attribution and revealing hidden links between seemingly unrelated accounts.