Blog

Security

From Recon to Free Flights: Precision Prompt Attacks on AI Agents

Guy Amit is a Senior AI Security Researcher at Akamai and holds a PhD from Ben-Gurion University, with a specialization in AI security and privacy. His work focuses on identifying emerging threats to generative AI systems, large language models, and agentic applications. His research spans offensive and defensive AI security, including prompt-based attacks, model vulnerabilities, and runtime protection techniques. Guy combines academic research with practical security work to develop new methods for evaluating and strengthening the resilience of AI-powered systems.

Key takeaways

Successful prompt attacks against AI agents rarely rely on generic phrases. Instead, they mimic classic cyberattacks by systematically probing the system to map tool surfaces, extract system prompt rules, and learn data schemas.

Attackers can inject fabricated tool responses and fake assistant messages directly into the conversation history.…