Back
Jassi Pannu Congressional Testimony
governmentCredibility Rating
5/5
Gold(5)Gold standard. Rigorous peer review, high editorial standards, and strong institutional reputation.
Rating inherited from publication venue: US Congress
Expert congressional testimony from a Johns Hopkins biosecurity researcher and former Google AI practitioner, offering concrete 2025 legislative recommendations on AI-biology dual-use risks; relevant to AI governance and catastrophic risk policy discussions.
Metadata
Importance: 62/100guidance documentprimary source
Summary
Dr. Jassi Pannu testifies before the House Committee on Energy and Commerce on biosecurity risks from AI applied to biology, providing five concrete legislative recommendations. She emphasizes targeted risk assessment for high-risk biological AI models, tracking LLMs and autonomous systems, formalizing biological data controls, modernizing NIH biosafety oversight for AI-designed agents, and legislating de novo gene synthesis. The testimony frames proactive biosecurity governance as both life-saving and economically beneficial.
Key Points
- •Recommends granting CAISI authority to define and regulate the narrow subset of biological AI models posing societal-scale risks.
- •Calls for resourcing CAISI to track benefits and risks of LLMs, AI agents, and autonomous robotics in biology via benchmarks and evaluations.
- •Urges OSTP to formalize a targeted framework for biological data controls to prevent misuse of sensitive biological information.
- •Highlights need to modernize NIH biosafety oversight to address AI-designed biological agents and toxins.
- •Advocates for legislation governing de novo gene synthesis as a critical biosecurity gap at the AI-biology frontier.
Cited by 1 page
| Page | Type | Quality |
|---|---|---|
| Johns Hopkins Center for Health Security | Organization | 63.0 |
Cached Content Preview
HTTP 200Fetched Mar 15, 202637 KB
Testimony of
Jassi Pannu, MD
Senior Scholar, Johns Hopkins Center for Health Security
Assistant Professor, Johns Hopkins Bloomberg School of Public Health
Before the
US House of Representatives Committee on Energy and Commerce
Subcommittee on Oversight and Investigations
“Examining Biosecurity at the Intersection of AI and Biology”
December 17, 2025
1
-- 1 of 13 --
Introduction
Chairman Joyce, Ranking Member Clarke, Chairman Guthrie, Ranking Member Pallone, and distinguished
members of the Subcommittee, thank you for the opportunity to appear before you today to discuss
biosecurity issues at the intersection of artificial intelligence (AI) and biology. The application of AI to
solving challenging health and biomedical problems could be one of the most beneficial uses of this
technology for American society and the world. In order to harness these benefits, we should seek to
understand, anticipate, and prevent, the narrow set of biosafety and biosecurity risks that could significantly
impact society. I commend the subcommittee for their attention to these issues.
I am currently an Assistant Professor at the Johns Hopkins Bloomberg School of Public Health, and a Senior
Scholar at the Johns Hopkins Center for Health Security. The opinions expressed herein are my own and do
not necessarily reflect the views of Johns Hopkins University.
I am a medical doctor by training, and completed my medical degree and internal medicine residency at
Stanford University. During the peak of COVID-19 pandemic, I worked full time on the frontlines caring for
patients, and I have worked as a doctor internationally in Uganda, a country that frequently faces both
endemic and emerging infectious disease outbreaks. I also previously worked at Google AI, where I
developed AI-enabled diagnostics. These experiences have shaped my view that technology advances can
provide immense public health benefits; but also, that the prevention of biothreats with societal impact is
essential. Prevention saves lives. It is also cost-effective and protects the American economy.
My work now focuses on tracking AI advancements for biology in order to develop technology and policy
approaches for mitigating risks. I also investigate ways in which AI could improve our public health and
biodefense readiness.
Today, I was asked to provide comments on how we can guard against potential harms of AI while at the
same time working to ensure that AI will improve the nation’s biosecurity and public health outcomes. Prior
to offering these comments, I want to share my top-line recommendations as to the actions Congress should
consider to address the biosafety and biosecurity risks of AI.
To that end, I recommend that Congress:
1) Provide CAISI with the authority to define the narrow subset of biological AI models that pose
societal risks, to enable targeted risk assessment and mitigation;
2) Resource CAISI to closely track the benefits and risks of LLMs, AI agents, and autonomous robotics
for biology,
... (truncated, 37 KB total)Resource ID:
518a2ccfe835dc2f | Stable ID: OWJjZDhjY2