Enhancing Botnet Detection with AI using LLMs and Similarity Search


As botnets continue to evolve, so do the techniques required to detect them. While Transport Layer Security (TLS) encryption is widely adopted for secure communications, botnets leverage TLS to obscure command-and-control (C2) traffic. These malicious actors often have identifiable characteristics embedded within their TLS certificates, opening a potential pathway for advanced detection techniques.

In first-of-its-kind research, Rapid7’s Dr. Stuart Millar, in collaboration with Kumar Shashwat, Francis Hahn and Prof. Xinming Ou, at the University of South Florida, studied the use of AI large language models (LLMs) to detect botnets’ use of TLS encryption by analyzing embedding similarities to weed out botnets within a sea of benign TLS certificates.

Read more…
Source: Rapid7


Sign up for our Newsletter


Related:

  • Satellites found exposing unencrypted data, including phone calls and some military comms

    October 14, 2025

    Security researchers have discovered that as many as half of all geostationary satellites in Earth’s orbit are carrying unencrypted sensitive consumer, corporate, and military information, making this data wide open to eavesdropping. The researchers at UC San Diego and the University of Maryland spent $800 on an off-the-shelf satellite receiver and pointed it at the sky ...

  • Qantas says customer data released by cyber criminals months after cyber breach

    October 12, 2025

    Australia’s Qantas Airways said on Sunday that it was one of the companies whose customer data had been published by cybercriminals after it was stolen by a hacker in a July breach of a database containing the personal information of the airline’s customers. The airline said in July that more than a million customers had sensitive ...

  • ClayRat Android malware spoofs WhatsApp, TikTok and more

    October 10, 2025

    A new Android malware variant is posing as popular apps, stealing sensitive files and propagating further. Experts from Zimperium revealed ClayRat, targeting primarily Russian users by spoofing popular Android apps such as WhatsApp, TikTok, Google Photos, or YouTube, distributed mostly through Telegram channels and standalone phishing sites. Through typosquatting, the phishing sites trick victims into thinking ...

  • The Golden Scale: Bling Libra and the Evolving Extortion Economy

    October 10, 2025

    In recent months, threat actors claiming to be part of a new conglomerate dubbed Scattered Lapsus$ Hunters (aka ​​SP1D3R HUNTERS, SLSH) have asserted responsibility for laying siege to customer Salesforce tenants as part of a coordinated effort to steal data and hold it for ransom. At least one industry source refers to this criminal syndicate as ...

  • SonicWall confirms all of its cloud backup customers were affected by data breach

    October 10, 2025

    All companies using SonicWall’s MySonicWall cloud backup feature have had their firewall configuration files exposed in a recent cyberattack, the company has admitted. After initially claiming “fewer than 5%” of its customer base was affected, the company has revealed the true scale of the incident. In mid-September 2025, SonicWall warned its firewall customers to reset their ...

  • Identifying and Mitigating Potential Velociraptor Abuse

    October 9, 2025

    Open-source technologies and communities are a big part of the Rapid7 ethos, and that’s not by chance – it’s by design. Rapid7 believe that their Metasploit, AttackerKB, and Velociraptor initiatives help create a strong threat intelligence foundation as well as a secure digital future for all. Unfortunately, the same open-source tools that help security teams ...