Planting the Honey SEAD: Active Phishing Defense with AI
This project is based on an RSA 2025 talk submission of the same name. If the talk is accepted the students will be credited for their help with the project. What follows is an abbreviated version of the actual submission.
This presentation describes a social engineering honeypot to lure online scammers to engage with fake employees. AI-powered agents posed as employees, responding to all contact attempts, with the intention of consuming threat actor time, energy, and resources. A common misperception in security is that defense is reactive, and therefore always behind the attackers. We destroy this misperception by demonstrating the employment of AI-powered cognitive agents within a defense portfolio. We will provide the audience with a high-level overview of how to build and implement their own “SEAD Bots” to frustrate attacker efforts.
Objectives
- Students will create a honeypot website with the intent of receiving phishing messages from scammers.
- Students will create an automated AI powered system to respond to these messages with the intent of maximizing scammer effort and engagement.
- Students will record their development process and measure success statistics to create a final report summarizing the process and its effectiveness.
- If the talk is accepted, the project will be presented by myself, Isaac Hathaway, at the RSA conference in April 2025.
Motivations
This project seeks to utilize AI-powered agents to combat phishing attacks by consuming attacker resources. Typically when a scammer receives a response they know that they are talking to a potential victim, but by using AI to respond to all of their phishing attacks we seek to trick them into spending resources on attacking non-targets. The end goal is to achieve a form of target hardening where phishing attacks are no longer viable, thus causing threat actors to give up.
Qualifications
Minimum Qualifications:
- Experience with Python
- Experience with Web Development
- Experience with AI Integration
Details
Project Partner:
Isaac Hathaway
NDA/IPA:No Agreement Required
Number Groups:1
Project Status:Accepting Applicants
Website:https://hathaway.llc/
Keywords:
SecurityArtificial Intelligence AILLMPhishing