Lubomir Lou Chitkushev*
Boston University, Boston, USA
ltc [at] bu.edu
Abstract
Cyberbiosecurity threats are rapidly evolving at the intersection of artificial intelligence, biotechnology, and digital infrastructure. As AI systems become increasingly embedded in biological research, healthcare, agriculture, and bio-manufacturing, they introduce new classes of vulnerabilities that extend beyond data breaches to real-world biological consequences. This talk examines how AI-enabled tools—ranging from generative models for protein design to automated laboratory platforms—expand the attack surface for malicious actors, enabling risks such as the manipulation of genomic data, the corruption of biofoundry workflows, and the design or synthesis of harmful biological agents.
We explore emerging threat scenarios, including adversarial attacks on biological datasets, exploitation of cloud-connected lab equipment, and the misuse of AI models trained on sensitive biological information. Particular attention is given to the convergence of cyber and physical systems, where compromised digital inputs can directly alter experimental outcomes or production pipelines. The talk also assesses current gaps in governance, security practices, and cross-disciplinary awareness that leave critical systems exposed.
Finally, we outline a framework for mitigating cyberbiosecurity risks in the age of AI, emphasizing secure-by-design principles, robust access controls, model auditing, and collaboration between cybersecurity experts and life scientists. By highlighting both technical and policy dimensions, this talk aims to inform a proactive approach to safeguarding biological systems against digitally mediated threats.
Keywords: Cyberbiosecurity, mitigating cyberbiosecurity risk