Fred Heiding Personal Site

fredheiding.com · 375 words · saved by 1 readers

Large language models are becoming exceedingly sophisticated at deceiving users. We continuously evaluate their capability to conduct personalized phishing attacks and compare their results with human experts and similar phishing studies conducted using AI models developed in previous years. Currently (2024), AI-automated attacks performed on par with human experts and 350% better than the control group. The results are a significant improvement from similar studies conducted last year (2023), highlighting the rapidly improving deceptive capabilities of AI models. We use a custom-built AI-based phishing tool that automates the entire phishing process, including information gathering and the creation of personalized vulnerability profiles of each target. We use the collected information to create personalized spam filters that adapt to each user's unique needs. Overview of AI-automated phishing campaigns. The process includes target identification, synthetic attacker profile creation, p

Postdoc at Harvard researching computer security and AI security with Eric Rosenbach and Bruce Schneier. Expert in AI-assisted cyberattacks, phishing, and national cybersecurity strategies. Publications and research projects on AI security, cybersecurity, and phishing attacks by Fred Heiding, postdoctoral researcher at Harvard. AI security projects and cybersecurity tools developed by Fred Heiding at Harvard, including research on automated phishing detection and AI-assisted cyberattacks. Courses and teaching experience by Fred Heiding, including AI security, cybersecurity, and generative AI i

Fred Heiding Personal Site

related reading