Artificial intelligence and the future of biodefense readiness

By Jassi Pannu

This is a lightly edited excerpt of testimony recently provided to the U.S. House’s Energy and Commerce Oversight and Investigations Subcommittee hearing titled, “Examining Biosecurity at the Intersection of AI and Biology.”

Within the biological sciences, AI systems will likely provide immense benefit. They are likely to be employed to improve the discovery, diagnosis and treatment of diseases, to boost agricultural yields, and to optimize the biosynthesis of useful products, among many other uses now being explored. AI also holds great promise in areas of biosecurity and biodefense. However, these areas are under-incentivized by the private market, and companies may not seek to apply AI for the improvement of biosecurity and biodefense by default. The government should consider investing in defense-forward applications of AI in order to strengthen our resilience to biological threats.  

Bio surveillance and diagnostics 

AI should be employed to improve our ability to detect outbreaks early by advancing pathogen bio-surveillance and scalable diagnostics. For example, AI could be used to synthesize multiple data streams, such as symptom search queries, wastewater sequence data, and vital signs from wearable devices.

Data generation 

A current barrier to biological AI model capabilities is high-quality data. A growing proportion of wet-lab work for data generation, such as data visualization, data analysis, and sample creation, can be conducted by autonomous machines, including machines that researchers pay to access remotely, known as cloud labs; this work could be further scaled using AI.  

Medical countermeasure development, production and distribution 

Within medical countermeasure production and distribution, AI can potentially: (1) increase supply chain optimization through demand forecasting, inventory management, production capacity planning and distribution network optimization; (2) enhance strategic national stockpile management through real-time tracking, predictive maintenance and automated procurement; and (3) improve manufacturing coordination through production line scheduling, cross-manufacturer standardization and quality assurance.  

Materials and hardware development and use 

Current biothreat readiness is limited by the types of available personal protective equipment and built environment pathogen disinfection technologies. AI could be used to advance the materials science behind filters for reusable respirators and to help users correctly fit test respirators themselves.  

Strengthening health systems' resilience  

Within healthcare personnel and resource deployment, AI can strengthen healthcare workforce management through: (1) personnel availability tracking, skills-based deployment matching, training coordination and burnout prediction; (2) optimizing healthcare facility coordination through bed capacity management, equipment allocation and surge capacity planning; and (3) enhance emergency medical services coordination through ambulance dispatch optimization, emergency department capacity management and inter-facility patient transport.  

Response coordination and treaty monitoring 

AI agents could be leveraged for rapid outbreak response coordination, and LLMs could be used to disseminate reliable public health information. AI tools for open-source intelligence could be leveraged to, for example, ensure compliance with treaties such as the Biological Weapons Convention.  

Read the full testimony here.

Jassi Pannu is an assistant professor at Johns Hopkins Bloomberg School of Public Health.

_{*The opinions expressed in this column are those of the author and do not necessarily reflect the views of HealthPlatform.News.}