Boyan Milanov is currently a Senior Research Scientist evaluating the cyber security risks associated with the integration of generative AI agents and components in the context of national security, defense, and safety-critical infrastructure. He was previously a senior security engineer at Trail of Bits where he focused on the unique security and safety challenges posed by emerging AI systems and capabilities. He worked with some of the world's top cybersecurity organizations, including DARPA, UK's National Cyber Security Center (NCSC), and UK's AI Safety Institute (AISI), helping them study, understand and address novel AI security threats.
Boyan has contributed to exposing new attacks affecting the AI supply chain and led the development of defensive tooling against these attack vectors. This led to his involvement in AI Bill of Material (AIBOM) efforts toward traceability and transparency of AI systems and data. Boyan also has an extensive background in designing and executing evaluations of frontier AI models across a wide range of cyber-capabilities that can threaten national security and institutions. He has also conducted cutting-edge research on State-Of-The-Art offensive AI systems as part of DARPA's AI Cyber Challenge.
Previously, Boyan also worked as a cybersecurity expert at Quarkslab where he was responsible for leading software security audits, delivering reverse engineering training, and conducting research on key security areas such as binary exploitation, symbolic execution, program instrumentation and tracing.
Boyan has contributed to exposing new attacks affecting the AI supply chain and led the development of defensive tooling against these attack vectors. This led to his involvement in AI Bill of Material (AIBOM) efforts toward traceability and transparency of AI systems and data. Boyan also has an extensive background in designing and executing evaluations of frontier AI models across a wide range of cyber-capabilities that can threaten national security and institutions. He has also conducted cutting-edge research on State-Of-The-Art offensive AI systems as part of DARPA's AI Cyber Challenge.
Previously, Boyan also worked as a cybersecurity expert at Quarkslab where he was responsible for leading software security audits, delivering reverse engineering training, and conducting research on key security areas such as binary exploitation, symbolic execution, program instrumentation and tracing.