---
title: "How AI Is Lowering Barriers to Biological Weapons and What Defenses Are Needed"
id: "stat-news-1-opinion-ai-is-eroding-the-barriers-that-kept-biological-weapons-rare"
canonical_url: "https://medichelpline.com/clinical-feed/stat-news-1-opinion-ai-is-eroding-the-barriers-that-kept-biological-weapons-rare"
content_type: "clinical_feed_article"
specialty: "Infectious Disease"
source_name: "STAT News"
source_url: "https://www.statnews.com/2026/09/28/ai-bioweapons-pathogens-guardrails-policy-warning/?utm_campaign=rss"
published_at: "2026-09-28T08:30:00.000Z"
evidence_level: "Verified Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# How AI Is Lowering Barriers to Biological Weapons and What Defenses Are Needed
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/stat-news-1-opinion-ai-is-eroding-the-barriers-that-kept-biological-weapons-rare
- **Specialty:** [Infectious Disease](https://medichelpline.com/clinical-feed/infectious-disease.md)
- **Primary Source:** STAT News
- **Source URL:** [Original Journal Publication](https://www.statnews.com/2026/09/28/ai-bioweapons-pathogens-guardrails-policy-warning/?utm_campaign=rss)
- **Published At:** 2026-09-28T08:30:00.000Z
- **Evidence Rating:** Verified Feed
## Executive GIST (TL;DR)
- AI models trained on biological literature can connect expertise across subfields, enabling individuals with limited backgrounds to perform steps previously requiring multidisciplinary teams. This narrows historical barriers that kept **biological weapons** rare. - Anthropic reported five instances where its models were used for work that could support bioweapons development, including gain-of-function proposals and experiments to increase cross-species infectivity; the company blocked those uses despite ambiguous intent. - Open-weight models with downloadable code often ship with fewer safeguards and limited monitoring compared with frontier labs that implement guardrails such as Anthropic and OpenAI. - Independent reports from Harvard and RAND concluded that AI could broaden the set of actors capable of mounting large-scale biological attacks and could make state programs more capable; neither report claims the most dangerous thresholds are already crossed but warns the risk is increasing. - Historical controls on biological weapons rested on the difficulty of combining skills across immunology, virology, cultivation, and stabilization; AI is eroding that gap by synthesizing cross-disciplinary guidance. - A Nature Communications paper showed 36 of 38 DNA synthesis companies supplied fragments of the 1918 influenza virus when ordered, illustrating weaknesses in current screening and the challenge of regulating shorter, reassemblable fragments. - The U.S. Select Agent Program prohibits intact genetic material capable of producing dangerous pathogens, but shorter DNA fragments remain largely unregulated and the DNA synthesis market is global, so blocked orders in one jurisdiction can be filled elsewhere. - Even with perfect DNA screening, access to materials globally and the ability of AI to create rapid genetic blueprints mean a pathogen design could be converted into a physical agent in weeks, although technical failure rates remain nontrivial. - Broad restrictions on biological research risk slowing beneficial biomedical advances; therefore the author argues for layered defenses rather than blanket bans. - Recommended layers include controlled access to advanced biological models with oversight, pre-release testing of models for dangerous capabilities, verification and screening by DNA synthesis providers including checks for split orders and third-party audits, and pushing open-weight developers toward similar testing standards while recognizing limited enforcement. - Slowing model development (“pace the frontier”) can buy time only if that interval is used to build biodefense; defenses take years to construct and cannot be deferred. - The author emphasizes urgency: defenses and guardrails must be built now because the threat landscape is advancing faster than current protections.
## Clinical Analysis & Structured Key Points
[Skip to Main Content](https://www.statnews.com/2026/09/28/ai-bioweapons-pathogens-guardrails-policy-warning/?utm_campaign=rss#wrapper) [ ![STAT](https://www.statnews.com/wp-content/themes/stat/images/stat-logo-white.svg) ![STAT](https://www.statnews.com/wp-content/themes/stat/images/stat-logo.svg) ](https://www.statnews.com/) * [Manage alerts for this article](https://www.statnews.com/my-account/edit/#emails) * [Save this story for later](https://www.statnews.com/my-account/saved/) * [Share this article](https://www.statnews.com/2026/09/28/ai-bioweapons-pathogens-guardrails-policy-warning/?utm_campaign=rss) [FDA](https://www.statnews.com/topic/fda/) [Congress](https://www.statnews.com/topic/congress/) [Vaccines](https://www.statnews.com/topic/vaccines/) * [Newsletters](https://www.statnews.com/signup/) * [Log In](https://www.statnews.com/login/) * [My Account](https://www.statnews.com/my-account/) * [Subscribe Now](https://www.statnews.com/stat-plus/) My Account * [Settings Complete your personal information for a more tailored experience ](https://www.statnews.com/my-account/) * [Saved Stories](https://www.statnews.com/my-account/saved/) * [Billing](https://www.statnews.com/my-account/billing/) * [Log In](https://www.statnews.com/login/) [ ![STAT](https://www.statnews.com/wp-content/themes/stat/images/stat-logo.svg) Reporting from the frontiers of health and medicine ](https://www.statnews.com/) * [Newsletters](https://www.statnews.com/signup/) * [Log In](https://www.statnews.com/login/) * [My Account](https://www.statnews.com/my-account/) * [Subscribe Now](https://www.statnews.com/stat-plus/) * [Biotech](https://www.statnews.com/category/biotech/) * [Pharma](https://www.statnews.com/category/pharma/) * [Public Health](https://www.statnews.com/category/health/) * [Health Tech](https://www.statnews.com/category/health-tech/) * [Policy](https://www.statnews.com/category/politics/) * [Science](https://www.statnews.com/category/in-the-lab/) * [First Opinion](https://www.statnews.com/category/first-opinion/) * [STAT+](https://www.statnews.com/topic/stat-plus/) * [STAT Events](https://www.statnews.com/events/) [ ![STAT](https://www.statnews.com/wp-content/themes/stat/images/stat-logo.svg) ](https://www.statnews.com/) Search * [Log In](https://www.statnews.com/login/) * [Subscribe Now](https://www.statnews.com/stat-plus/) * [My Account](https://www.statnews.com/my-account/) My Account * [Settings](https://www.statnews.com/my-account/) * [Saved Stories](https://www.statnews.com/my-account/saved/) * [Billing History](https://www.statnews.com/my-account/billing/) [Home](https://www.statnews.com/) News * [Latest](https://www.statnews.com/latest/) * [Series & Investigations](https://www.statnews.com/stat-series-and-investigations-award-winning-journalism/) * Business * [Biotech](https://www.statnews.com/category/biotech/) * [Pharma](https://www.statnews.com/category/pharma/) * [Health Tech](https://www.statnews.com/category/health-tech/) * [Health Insurance](https://www.statnews.com/category/insurance/) * [Hospitals](https://www.statnews.com/category/hospitals/) * [Medical Devices](https://www.statnews.com/topic/medical-devices/) * Washington * [Policy](https://www.statnews.com/category/politics/) * [FDA](https://www.statnews.com/topic/fda/) * [NIH](https://www.statnews.com/topic/nih/) * Science * [CRISPR](https://www.statnews.com/topic/crispr/) * [Gene Therapy](https://www.statnews.com/topic/gene-therapy/) * [Neuroscience](https://www.statnews.com/topic/neuroscience/) * Public Health * [Addiction](https://www.statnews.com/topic/addiction/) * [Covid-19](https://www.statnews.com/topic/coronavirus/) * [Abortion](https://www.statnews.com/topic/abortion/) * [Health Disparities](https://www.statnews.com/topic/health-disparities/) * [Infectious Disease](https://www.statnews.com/topic/infectious-disease/) * [Mental Health](https://www.statnews.com/topic/mental-health/) * Disease * [Cancer](https://www.statnews.com/topic/cancer/) * [Cardiovascular Disease](https://www.statnews.com/topic/cardiovascular-disease/) * [Chronic Disease](https://www.statnews.com/topic/chronic-disease/) * [Diabetes](https://www.statnews.com/topic/diabetes/) * [Alzheimer's](https://www.statnews.com/topic/alzheimers/) * [Obesity](https://www.statnews.com/topic/obesity/) * Features * [Health Care Costs](https://www.statnews.com/health-insurance-costs-out-of-pocket-out-of-reach-series/) * [The Deadliest Drug](https://www.statnews.com/alcohol-epidemic-hiding-plain-sight-deadliest-drug-series/) * [Science, Shattered](https://www.statnews.com/american-science-shattered/) [Newsletters](https://www.statnews.com/signup/) [Opinion](https://www.statnews.com/category/first-opinion/) Columns * [Adam Feuerstein](https://www.statnews.com/category/adams-take/) * [Matthew Herper](https://www.statnews.com/category/matts-take/) * [Ed Silverman](https://www.statnews.com/category/pharmalot-view/) [Reports](https://reports.statnews.com/) [E-books](https://www.statnews.com/ebooks/) [Podcasts](https://www.statnews.com/podcasts/) [Tools & Trackers](https://www.statnews.com/trackers/) * [CRISPR Tracker](https://www.statnews.com/feature/crispr/tracker/) * [Breakthrough Device Tracker](https://www.statnews.com/feature/stat-plus/breakthrough-device-designation-fda-tracker/) * [Generative AI Tracker](https://www.statnews.com/feature/stat-plus/generative-ai-health-hospitals-tracker/) * [Obesity Drug Tracker](https://www.statnews.com/2023/09/12/new-weight-loss-drug-tracker-novo-nordisk-eli-lilly/) [Events](https://www.statnews.com/events/) * [Upcoming Events](https://www.statnews.com/events#upcoming-events) * [Summits](https://www.statnews.com/stat-summits/) [Community](https://www.statnews.com/community/) * [STAT Wunderkinds 2026](https://www.statnews.com/wunderkinds-nominate/) [Video](https://youtube.com/@Statnews) [Advertise](https://www.statnews.com/advertise/) * [STAT Brand Studio](https://sbs.statnews.com/) What's the word? Test your knowledge with our new weekday mini crossword [ Start solving ](https://www.statnews.com/stat-mini-crossword/) Opinion[First Opinion](https://www.statnews.com/category/first-opinion/) # AI is eroding the barriers that kept biological weapons rare ## A genetic blueprint created by AI can become an actual pathogen in weeks * [Manage alerts for this article](https://www.statnews.com/my-account/edit/#emails) * [Save this story for later](https://www.statnews.com/my-account/saved/) * [Share this article](https://www.statnews.com/2026/09/28/ai-bioweapons-pathogens-guardrails-policy-warning/?utm_campaign=rss) ![](https://www.statnews.com/wp-content/uploads/2026/09/AdobeStock_646357491-645x645.jpeg) Adobe By Ashish K. Jha Sept. 28, 2026 Jha, a former White House Covid-19 response coordinator, is a senior fellow at the Belfer Center at Harvard Kennedy School. He is also co-founder and CEO of BioRadar, a public benefit company that builds systems to detect biological threats. Earlier this month, [Anthropic reported five cases](https://www.anthropic.com/threat-intelligence-report-september-2026) in which people used its models for work that could support biological weapons development. The cases included help preparing a proposal for gain-of-function research on a mosquito-borne virus at a military institute and planning experiments intended to help bird flu get better at infecting mammals. Anthropic had no way to determine the intent behind this work but blocked it anyway. That ambiguity is the problem. The same tools that scientists might use to design a vaccine can help design a worse pathogen. The same tools that help scientists design a drug that targets cancer cells while sparing healthy ones can be tweaked to build a drug that targets healthy cells alone. There is no need for someone to ask Claude to “build a bioweapon.” They only need to look like scientists who are working on important biological problems. Advertisement Anthropic, which shut the accounts down, has been the most public about the risks of AI to the development of bioweapons. OpenAI has also put in guardrails. Other AI companies, especially with open-weight models — those whose underlying code anyone can download and modify without any company oversight — generally release their products with far fewer safeguards and little monitoring for how their tools are being used. Much of the industry’s safety debate is about whether to “pace the frontier” — slowing model development so guardrails and defenses can catch up. That’s worth doing, but it only works if it is a bridge to building biodefense. And slowing model development certainly has no impact on who can order synthetic DNA —or what national governments are doing to build up their capabilities. For decades, biological weapons were rare for a reason: Building them takes expertise that’s hard to acquire and harder to combine. Immunology, virology, growing a pathogen, and stabilizing a dangerous mutation all require different skills. The people who could connect those pieces are few and far between. Advertisement Related Story ![](https://www.statnews.com/wp-content/uploads/2025/05/GettyImages-1210642602-768x432.jpg) ### [The AI threat to public health no one is thinking about: a fake bioterrorist attack](https://www.statnews.com/2025/05/27/artificial-intelligence-bioterrorism-deepfake-public-health-threat/) AI is closing those gaps. Models trained on the biological literature can connect ideas across subfields that few researchers have mastered alone. These models can make someone with relevant training in one area more capable and dangerous by helping them learn key steps in another area. Two reports this year — one from [Harvard](https://www.belfercenter.org/bio-deterrence), one from [RAND](https://www.axios.com/2026/08/21/safeguards-ai-bioweapons-roadmap) — examined these issues and reached similar conclusions independently: AI could broaden the range of actors able to mount a large-scale biological attack, while making existing state programs more capable, too. Neither claims the most dangerous thresholds have been crossed though neither rules it out. The safer assumption is that we’re already close enough not to wait and find out. The [State Department’s annual declassified intelligence report](https://councilonstrategicrisks.org/2025/04/21/the-state-of-compliance-with-weapons-of-mass-destruction-related-treaties-2) has repeatedly concluded that Russia and North Korea have offensive biological weapons programs, while raising concerns about China and Iran. In 2024, [The Washington Post documented a major expansion at Sergiev Posad-6](https://www.washingtonpost.com/national-security/interactive/2024/russia-biological-chemical-weapons-laboratory-expansion/), a former Soviet bioweapons site outside Moscow, including new high-containment buildings. Most of the activity that the declassified intelligence reports point to predates AI. With AI, state programs can expand the number of novel designs. Non-state actors — terrorist groups and others — have also tried to get access to these weapons by building capabilities themselves or attacking facilities that already have the capability. We are talking not about a kid in a basement building the weapon but about serious actors with resources, enabled by biological engineering tools and fueled by AI. And then, there is the second barrier — which is also eroding quickly. A paper published this year in [Nature Communications](https://www.nature.com/articles/s41467-025-67955-3) described an experiment in which researchers ordered genetic fragments of the 1918 influenza virus and 36 of the 38 DNA synthesis companies sent the sequences. It was enough for a skilled person to reconstruct the virus. While the [U.S. Select Agent Program](https://www.selectagents.gov/), a federal program overseen by the CDC and th Department of Agriculture, prohibits intact genetic material capable of producing dangerous pathogens, shorter, readily assembled fragments remain unregulated. Advertisement But here is the bigger point: Even if we had a perfect screening system that prevented people from ordering DNA fragments that could later be reassembled, it would not be enough. DNA synthesis is a global market, and a sequence blocked in one country can be ordered from another. Someone, somewhere, will get access to the materials they want if they want it badly enough. The threat landscape is moving faster than our defenses. A genetic blueprint created by AI can become an actual pathogen in weeks. Yes, it is still hard to turn an idea into an actual pathogen — the failure rate is high — but enough actors are trying that some of them will succeed. Biomedicine is entering one of its most productive periods in decades, driven substantially by these same tools. Broad restrictions would slow research into diseases we could otherwise understand and treat faster. There is no reasonable way to put in restrictions to prevent the use of tools to build bioweapons without thwarting progress in the next generation of therapies. So what else to do? We need a layered defense, which begins with access and model design. It’s reasonable for verified researchers to have access to leading-edge biological models for legitimate work, but appropriate oversight must be required. And the developers should test models for dangerous capabilities before release. Governments could require DNA synthesis providers to verify customers, screen for split orders, and undergo independent third-party audits to make sure the system is working. While there is broad consensus that this should be done, whether it will be particularly effective is not clear. Related Story ![](https://www.statnews.com/wp-content/uploads/2024/11/GettyImages-2032109653-768x432.jpg) ### ![](https://www.statnews.com/wp-content/themes/stat/images/home/statplus.svg) [STAT Plus: AI titans push Congress for DNA safeguards](https://www.statnews.com/2026/06/04/biotech-news-ai-titans-push-congress-to-dna-safeguards/) Similarly, getting open-weight AI developers to adopt that same testing standard and guardrails as the frontier AI labs (such as Anthropic and OpenAI) is worth pushing for, but expecting full compliance is unrealistic. And it’s worth remembering that those models, once released, can’t be recalled. Much of the AI regulation debate is about speed, autonomy, and who controls the frontier. A temporary slowdown might buy time on all three. Advertisement But that time is only worth buying if it goes toward building our defenses. We can’t depend on every lab, every country, and every open-source developer getting biosecurity right. They won’t. Over time, that means the development and potential use of biological weapons becomes more likely. Defenses against this threat take years to build. We don’t have years to spare. _Ashish K. Jha, a former White House Covid-19 response coordinator, is a senior fellow at the Belfer Center at Harvard Kennedy School. He is also co-founder and CEO of BioRadar, a public benefit company that builds systems to detect biological threats._ * * * ###### Letter to the editor Have an opinion on this essay? [Submit a letter to the editor](https://www.statnews.com/submit-a-letter-to-the-editor/?subject=AI%20is%20eroding%20the%20barriers%20that%20kept%20biological%20weapons%20rare). [Artificial intelligence](https://www.statnews.com/topic/artificial-intelligence/) [Submit a correction request](https://www.statnews.com/contact/?how_can_we_help_=Editorial%20Correction%20Request)[Reprints](https://www.parsintl.com/publications/stat/) 1. Ashish K. Jha 1. [](https://x.com/ashishkjha "X") 1. [](https://x.com/ashishkjha "X") ![](https://www.statnews.com/wp-content/themes/stat/images/article-new/envelope-open.svg) ![](https://www.statnews.com/wp-content/uploads/2023/06/First-Opinion-Logo-Black-640x319.png) ### Newsletter The smartest thinkers in life sciences on what's happening — and what's to come Please enter a valid email address. Sign Up Your data will be processed in accordance with our [Privacy Policy](https://www.statnews.com/privacy/) and [Terms of Service](https://www.statnews.com/terms-conditions/). You may opt out of receiving STAT communications at any time. Advertisement Most Popular 1. ![](https://www.statnews.com/wp-content/uploads/2021/09/AP20088463242938-80x80.jpg) ![](https://www.statnews.com/wp-content/themes/stat/images/home/statplus.svg) [Licensing deals on generic versions of Roche flu drug aimed at preparing for pandemic](https://www.statnews.com/pharmalot/2026/09/25/roche-xofluza-influenza-treatment-agreements/) 2. ![](https://www.statnews.com/wp-content/uploads/2025/08/GettyImages-2092473213-80x80.jpg) ![](https://www.statnews.com/wp-content/themes/stat/images/home/statplus.svg) [Lilly wins approval for once-weekly insulin shot](https://www.statnews.com/2026/09/25/biotech-news-lilly-wins-approval-for-weekly-insulin-shot/) 3. ![](https://www.statnews.com/wp-content/uploads/2021/03/Pharmalot_ComingGoing_STILL-80x80.jpg) ![](https://www.statnews.com/wp-content/themes/stat/images/home/statplus.svg) [Up and down the ladder: The latest comings and goings](https://www.statnews.com/pharmalot/2026/09/25/up-and-down-the-ladder-latest-comings-goings-jobs-pharma-biotech/) 4. ![](https://www.statnews.com/wp-content/uploads/2026/07/AdobeStock_1859455353-80x80.jpeg) ![](https://www.statnews.com/wp-content/themes/stat/images/home/statplus.svg) [Hospitals’ AI may be drifting. Who’s watching?](https://www.statnews.com/2026/07/22/hospital-ai-systems-ceo-leadership-accountability-governance/) 5. ![](https://www.statnews.com/wp-content/uploads/2026/02/epstein-80x80.jpg) ![](https://www.statnews.com/wp-content/themes/stat/images/home/statplus.svg) [Jeffrey Epstein’s tissue samples ignited a furor in the Harvard lab of George Church](https://www.statnews.com/2026/02/24/epstein-cell-line-george-church-harvard-personal-genome-project/) #### [ Reports ](https://reports.statnews.com/) [ ![](https://www.statnews.com/wp-content/uploads/2026/08/2026_CancerReport_ReportCover-328x424.jpg) Explore Cancer’s transformational moments: KRAS, CAR-T, immunotherapy, and more ](https://reports.statnews.com/products/cancer-transformational-moments-advances-in-kras-car-t-immunotherapy) Advertisement #### [Events](https://www.statnews.com/events/) [ ![](https://www.statnews.com/wp-content/uploads/2026/07/924_Samsung_VirtualEvent_Wordpress_FeaturedImage-80x80.jpg) Biosimilars: Have We Reached an Inflection Point? Virtual Sep 24 ](https://www.statnews.com/event/biosimilars-what-is-next-virtual-event-sept-2026/) [ ![](https://www.statnews.com/wp-content/uploads/2026/01/Flagship2026_Wordpress_CalendarImageRegGoLive-1-80x80.jpg) 2026 STAT Summit Boston, MA Oct 14 ](https://www.statnews.com/event/2026-stat-summit/) [ ![](https://www.statnews.com/wp-content/uploads/2026/07/2026_DD_OfficeHours_VirtualEvent_Wordpress_CalendarImage-80x80.jpg) Office Hours with Isabella Cueto and Lev Facher Virtual Jul 29 ](https://www.statnews.com/event/deadliest-drug-series-office-hours-ama-july-2026/) Most Read in [STAT+](https://www.statnews.com/stat-plus/) 1. ![](https://www.statnews.com/wp-content/uploads/2026/09/alkermes-900-winter-st-80x80.jpg) ![](https://www.statnews.com/wp-content/themes/stat/images/home/statplus.svg) [Alkermes’ orexin agonist shows potential to treat ADHD in early-stage trial](https://www.statnews.com/2026/09/21/alkermes-orexin-agonist-adhd-study/) 2. ![](https://www.statnews.com/wp-content/uploads/2026/09/AP6802385202387510-80x80.jpg) ![](https://www.statnews.com/wp-content/themes/stat/images/home/statplus.svg) [What a vote on Grail’s cancer detection test could mean for the company, patients, and regulators](https://www
## Related Clinical Research

- [Helen Branswell, veteran infectious disease reporter at STAT, dies at 69](https://medichelpline.com/clinical-feed/stat-news-0-helen-branswell-whose-reporting-on-infectious-diseases-and-public-health-drew.md)
- [Vaxcyte’s VAX-31 meets Phase 3 endpoints; FDA delays Novo Nordisk hemophilia A review over manufac](https://medichelpline.com/clinical-feed/stat-news-1-stat-pharmalittle-we-re-reading-about-a-vaxcyte-pneumococcal-shot-increasing.md)
- [Semaglutide plus menopause hormone therapy linked to greater weight loss in midlife women](https://medichelpline.com/clinical-feed/medical-news-today-0-ozempic-plus-hormone-therapy-may-help-boost-weight-loss-at-menopause.md)
- [High-Fiber Diet Outperforms Metformin in Blood Sugar Control and Weight Loss](https://medichelpline.com/clinical-feed/medical-news-today-0-high-fiber-diet-vs-metformin-which-is-better-for-blood-sugar-control-weight-loss.md)

## Navigation
- [← Back to Infectious Disease Feed](https://medichelpline.com/clinical-feed/infectious-disease.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.