Listen next week to hear from our Editor-in-Chief, Dr. Lassmann, about how our editorial team of infectious disease experts select & verify biothreat signals, then review, contextualize, edit, & publish the AI-drafted reports, allowing for more rapid reporting of new threats.
The hosts of @youmakemesickpod.bsky.social stopped by our offices today to chat with @bhadeliamd.bsky.social & Dr. Britta Lassman (@bumedicine.bsky.social) about infectious disease careers and CEID's latest work, including @beaconbio.bsky.social. Stay tuned to hear their episode next week!
February 12, 2026 at 7:50 PM
Listen next week to hear from our Editor-in-Chief, Dr. Lassmann, about how our editorial team of infectious disease experts select & verify biothreat signals, then review, contextualize, edit, & publish the AI-drafted reports, allowing for more rapid reporting of new threats.
A total of 1.25 million people died from tuberculosis (TB) in 2023. Worldwide, TB has probably returned to being the world's leading cause of death from a single infectious agent. (Source: WHO) #biosecurity #biothreat #onehealth
January 26, 2026 at 4:50 PM
A total of 1.25 million people died from tuberculosis (TB) in 2023. Worldwide, TB has probably returned to being the world's leading cause of death from a single infectious agent. (Source: WHO) #biosecurity #biothreat #onehealth
RAND proposed twelve measures to counter AI-enabled biothreat creation, including model unlearning, classifier-based input/output filtering for biological design tools, AI biosecurity agents, and enhanced biosurveillance.
Prioritizing Feasible and Impactful Actions to Enable Secure AI Development and Use in Biology
This paper provides policy and technology strategies collected from a diverse range of sources placed in the context of an organizing framework aligned with steps in the AI-enabled creation of a bioth...
www.rand.org
January 22, 2026 at 2:39 PM
RAND proposed twelve measures to counter AI-enabled biothreat creation, including model unlearning, classifier-based input/output filtering for biological design tools, AI biosecurity agents, and enhanced biosurveillance.
In this piece I’m quoted about the placeholder concept for an “unknown unknown” biothreat: Disease X. Though our technology is up to the task to give us resilience, the human factors - as exemplified by leaked - greatly diminish resiliency www.npr.org/2026/01/21/g...
Global buzzwords that will be buzzing in your ear in 2026
Will it be a year of "fractured resilience"? Or "pragmatic empathy"? Will "MOUs" be the next global health strategy? Are we in a new age of "decolonization" — or of "localization"?
www.npr.org
January 22, 2026 at 3:51 AM
In this piece I’m quoted about the placeholder concept for an “unknown unknown” biothreat: Disease X. Though our technology is up to the task to give us resilience, the human factors - as exemplified by leaked - greatly diminish resiliency www.npr.org/2026/01/21/g...
From Aerosol to Signal: Advances in Biosensor Technologies for Airborne Biothreat Detection https://pubmed.ncbi.nlm.nih.gov/41440245/
January 5, 2026 at 9:31 AM
From Aerosol to Signal: Advances in Biosensor Technologies for Airborne Biothreat Detection https://pubmed.ncbi.nlm.nih.gov/41440245/
From Aerosol to Signal: Advances in Biosensor Technologies for Airborne Biothreat Detection https://pubmed.ncbi.nlm.nih.gov/41440245/
December 29, 2025 at 8:30 AM
From Aerosol to Signal: Advances in Biosensor Technologies for Airborne Biothreat Detection https://pubmed.ncbi.nlm.nih.gov/41440245/
Weekly Recap: Writing better code without AI, AI in peer review, local LLMs, Biothreat Benchmark Generation, #Rstats updates (R Data Scientist, RWeekly, R Works), red-teaming an AI vending machine, #biosecurity, new papers doi.org/10.59350/t1q...
Weekly Recap (December 19, 2025)
Writing better code without AI, AI in peer review, local LLMs, Biothreat Benchmark Generation, R updates (R Data Scientist, RWeekly, R Works), red-teaming an AI vending machine, new papers
doi.org
December 19, 2025 at 10:53 AM
Weekly Recap: Writing better code without AI, AI in peer review, local LLMs, Biothreat Benchmark Generation, #Rstats updates (R Data Scientist, RWeekly, R Works), red-teaming an AI vending machine, #biosecurity, new papers doi.org/10.59350/t1q...
If one wanted to create a biothreat agent, B. burgdorferi is not the bug I'd start with. Difficult genome, difficult to grow, requires growth in a tick to become infectious in ways only partially understood. Space lasers would be easier to implement.
December 16, 2025 at 8:43 PM
If one wanted to create a biothreat agent, B. burgdorferi is not the bug I'd start with. Difficult genome, difficult to grow, requires growth in a tick to become infectious in ways only partially understood. Space lasers would be easier to implement.
Biothreat Benchmark Generation Framework for Evaluating Frontier AI Models
I: The Task-Query Architecture arxiv.org/abs/2512.08130
II: Benchmark Generation Process: arxiv.org/abs/2512.08451
III: Implementing the Bacterial Biothreat Benchmark (B3) Dataset: arxiv.org/abs/2512.08459
I: The Task-Query Architecture arxiv.org/abs/2512.08130
II: Benchmark Generation Process: arxiv.org/abs/2512.08451
III: Implementing the Bacterial Biothreat Benchmark (B3) Dataset: arxiv.org/abs/2512.08459
December 13, 2025 at 11:13 AM
Biothreat Benchmark Generation Framework for Evaluating Frontier AI Models
I: The Task-Query Architecture arxiv.org/abs/2512.08130
II: Benchmark Generation Process: arxiv.org/abs/2512.08451
III: Implementing the Bacterial Biothreat Benchmark (B3) Dataset: arxiv.org/abs/2512.08459
I: The Task-Query Architecture arxiv.org/abs/2512.08130
II: Benchmark Generation Process: arxiv.org/abs/2512.08451
III: Implementing the Bacterial Biothreat Benchmark (B3) Dataset: arxiv.org/abs/2512.08459
Gary Ackerman, et al.: Biothreat Benchmark Generation Framework for Evaluating Frontier AI Models III: Implementing the Bacterial Biothreat Benchmark (B3) Dataset https://arxiv.org/abs/2512.08459 https://arxiv.org/pdf/2512.08459 https://arxiv.org/html/2512.08459
December 10, 2025 at 6:33 AM
Gary Ackerman, et al.: Biothreat Benchmark Generation Framework for Evaluating Frontier AI Models III: Implementing the Bacterial Biothreat Benchmark (B3) Dataset https://arxiv.org/abs/2512.08459 https://arxiv.org/pdf/2512.08459 https://arxiv.org/html/2512.08459
Gary Ackerman, et al.: Biothreat Benchmark Generation Framework for Evaluating Frontier AI Models II: Benchmark Generation Process https://arxiv.org/abs/2512.08451 https://arxiv.org/pdf/2512.08451 https://arxiv.org/html/2512.08451
December 10, 2025 at 6:33 AM
Gary Ackerman, et al.: Biothreat Benchmark Generation Framework for Evaluating Frontier AI Models II: Benchmark Generation Process https://arxiv.org/abs/2512.08451 https://arxiv.org/pdf/2512.08451 https://arxiv.org/html/2512.08451
Gary Ackerman, et al.: Biothreat Benchmark Generation Framework for Evaluating Frontier AI Models I: The Task-Query Architecture https://arxiv.org/abs/2512.08130 https://arxiv.org/pdf/2512.08130 https://arxiv.org/html/2512.08130
December 10, 2025 at 6:33 AM
Gary Ackerman, et al.: Biothreat Benchmark Generation Framework for Evaluating Frontier AI Models I: The Task-Query Architecture https://arxiv.org/abs/2512.08130 https://arxiv.org/pdf/2512.08130 https://arxiv.org/html/2512.08130
4. Deep ignorance paper (August 2025) @kyletokens.bsky.social
We showed that filtering biothreat-related pretraining data is SOTA for making models resist adversarial fine-tuning. We proposed an amendment to the hypothesis from papers 1 and 2 above.
deepignorance.ai
We showed that filtering biothreat-related pretraining data is SOTA for making models resist adversarial fine-tuning. We proposed an amendment to the hypothesis from papers 1 and 2 above.
deepignorance.ai
November 25, 2025 at 8:00 PM
4. Deep ignorance paper (August 2025) @kyletokens.bsky.social
We showed that filtering biothreat-related pretraining data is SOTA for making models resist adversarial fine-tuning. We proposed an amendment to the hypothesis from papers 1 and 2 above.
deepignorance.ai
We showed that filtering biothreat-related pretraining data is SOTA for making models resist adversarial fine-tuning. We proposed an amendment to the hypothesis from papers 1 and 2 above.
deepignorance.ai
3. Estimating worst-case open-weight risks paper (Aug 2025)
They reported an instance where filtering biothreat data didn't have a big impact. But without more info on how and how much they filtered, it's hard to draw strong conclusions.
arxiv.org/abs/2508.03153
They reported an instance where filtering biothreat data didn't have a big impact. But without more info on how and how much they filtered, it's hard to draw strong conclusions.
arxiv.org/abs/2508.03153
November 25, 2025 at 8:00 PM
3. Estimating worst-case open-weight risks paper (Aug 2025)
They reported an instance where filtering biothreat data didn't have a big impact. But without more info on how and how much they filtered, it's hard to draw strong conclusions.
arxiv.org/abs/2508.03153
They reported an instance where filtering biothreat data didn't have a big impact. But without more info on how and how much they filtered, it's hard to draw strong conclusions.
arxiv.org/abs/2508.03153
Metabolomics ID's targets against the Burkholderia biothreat by finding metabolic shifts during infection. 🦠💊
pubmed.ncbi.nlm.nih.gov/41170848/
pubmed.ncbi.nlm.nih.gov/41170848/
Metabolomic profiling of Burkholderia thailandensis infection of airway epithelial cells provides insights into potential therapeutic targets - PubMed
<span><i>Burkholderia pseudomallei,</i> a gram-negative saprophyte that causes melioidosis, has been classified as a potential bioweapon, posing a serious global threat. Metabolic re-wiring, production of virulence effectors, evasion/suppression of host defenses, and modification of host cell functions co</span> …
pubmed.ncbi.nlm.nih.gov
November 25, 2025 at 7:18 PM
Metabolomics ID's targets against the Burkholderia biothreat by finding metabolic shifts during infection. 🦠💊
pubmed.ncbi.nlm.nih.gov/41170848/
pubmed.ncbi.nlm.nih.gov/41170848/
Jeff Bezos's AI Startup, DeepMind Weather Tech, OpenAI Biothreat Defense, NVIDIA Open Physics Models PLUS the latest AI tools, and more. Read the full newsletter here and subscribe to stay in the loop.
Jeff Bezos' AI Venture, DeepMind's Weather Tech, Cloudflare Acquisition, and NVIDIA's Simulation Tools
The RAIZOR Report: AI News, Tools & Jobs
raizor.co
November 19, 2025 at 1:30 AM
Jeff Bezos's AI Startup, DeepMind Weather Tech, OpenAI Biothreat Defense, NVIDIA Open Physics Models PLUS the latest AI tools, and more. Read the full newsletter here and subscribe to stay in the loop.
Metagenomic sequencing has been gaining traction, with The President’s FY 2026 Budget proposing a $52M allocation to CDC for Biothreat Radar, a new pathogen detection system.
November 12, 2025 at 6:30 PM
Metagenomic sequencing has been gaining traction, with The President’s FY 2026 Budget proposing a $52M allocation to CDC for Biothreat Radar, a new pathogen detection system.
I’m not going to say that mRNA vaccines are the be all and end all. But we have no other tool in our armamentarium that are as fast and as flexible an option as planning for a pandemic or other biothreat says @drsethberkley.bsky.social
October 30, 2025 at 8:51 PM
I’m not going to say that mRNA vaccines are the be all and end all. But we have no other tool in our armamentarium that are as fast and as flexible an option as planning for a pandemic or other biothreat says @drsethberkley.bsky.social
World Organisation for Animal Health hosts biothreat conference to address high stakes of global health security
www.woah.org/en/woah-host...
www.woah.org/en/woah-host...
World Organisation for Animal Health hosts biothreat conference to address high stakes of global health security - WOAH - World Organisation for Animal Health
In our increasingly connected and complex world, global health security is also a matter of national security, which makes veterinarians the frontline of defense.
www.woah.org
October 29, 2025 at 3:03 PM
World Organisation for Animal Health hosts biothreat conference to address high stakes of global health security
www.woah.org/en/woah-host...
www.woah.org/en/woah-host...
World Organisation for Animal Health hosts biothreat conference to address high stakes of global health security - WOAH
->World Organisation for Animal Health | #global health | More from Lil Dr Glen EcoChat at BigEarthData.ai
->World Organisation for Animal Health | #global health | More from Lil Dr Glen EcoChat at BigEarthData.ai
World Organisation for Animal Health hosts biothreat conference to address high stakes of global health security - WOAH
www.woah.org
October 27, 2025 at 1:29 PM
Disruptive biosecurity: A new concept for preparedness and response to the emerging total biothreat landscape www.tandfonline.com/doi/full/10.... #biosecurity 🧬🖥️🧪
October 10, 2025 at 5:55 PM
Disruptive biosecurity: A new concept for preparedness and response to the emerging total biothreat landscape www.tandfonline.com/doi/full/10.... #biosecurity 🧬🖥️🧪
“Biothreat hunters catch dangerous DNA before it gets made.”
We are super proud to be part of this red team effort with @erichorvitz.bsky.social & @msftresearch.bsky.social.
@silverjacket.bsky.social, @ibbis-bio.bsky.social, @jaimeyassif.bsky.social, @twistbioscience.com
We are super proud to be part of this red team effort with @erichorvitz.bsky.social & @msftresearch.bsky.social.
@silverjacket.bsky.social, @ibbis-bio.bsky.social, @jaimeyassif.bsky.social, @twistbioscience.com
Biothreat hunters catch dangerous DNA before it gets made
Microsoft researchers found a biological zero-day vulnerability — and created a software patch.
www.nature.com
October 8, 2025 at 2:08 PM
“Biothreat hunters catch dangerous DNA before it gets made.”
We are super proud to be part of this red team effort with @erichorvitz.bsky.social & @msftresearch.bsky.social.
@silverjacket.bsky.social, @ibbis-bio.bsky.social, @jaimeyassif.bsky.social, @twistbioscience.com
We are super proud to be part of this red team effort with @erichorvitz.bsky.social & @msftresearch.bsky.social.
@silverjacket.bsky.social, @ibbis-bio.bsky.social, @jaimeyassif.bsky.social, @twistbioscience.com
Biothreat hunters catch dangerous DNA before it gets made
On 2 October, an article in Nature says that DNA-synthesis firms routinely use biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. however, it warns that a paper published in Science describes…
On 2 October, an article in Nature says that DNA-synthesis firms routinely use biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. however, it warns that a paper published in Science describes…
Biothreat hunters catch dangerous DNA before it gets made
On 2 October, an article in Nature says that DNA-synthesis firms routinely use biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. however, it warns that a paper published in Science describes a potential vulnerability in this workflow. It says that the paper details how protein-design strategies aided by artificial intelligence (AI) could circumvent the screening software that many DNA-synthesis firms use to ensure that they avoid unintentionally producing sequences encoding harmful proteins or pathogens.
raytodd.blog
October 4, 2025 at 2:12 PM
Biothreat hunters catch dangerous DNA before it gets made
On 2 October, an article in Nature says that DNA-synthesis firms routinely use biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. however, it warns that a paper published in Science describes…
On 2 October, an article in Nature says that DNA-synthesis firms routinely use biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. however, it warns that a paper published in Science describes…
صيادون Biothreat يلتقطون الحمض النووي الخطير قبل أن يصنع
تستخدم شركات التخليق الحمض النووي بشكل روتيني-برامج فحص الأمن في الأمن لضمان عدم إنشاء تسلسلات خطيرة عن غير قصد. لكن ورقة نشرت في علوم في 2 أكتوبر يصف الضعف المحتمل في سير العمل هذا1. إنها تفاصيل كيف يمكن أن تحايل استراتيجيات تصميم البروتين…
تستخدم شركات التخليق الحمض النووي بشكل روتيني-برامج فحص الأمن في الأمن لضمان عدم إنشاء تسلسلات خطيرة عن غير قصد. لكن ورقة نشرت في علوم في 2 أكتوبر يصف الضعف المحتمل في سير العمل هذا1. إنها تفاصيل كيف يمكن أن تحايل استراتيجيات تصميم البروتين…
صيادون Biothreat يلتقطون الحمض النووي الخطير قبل أن يصنع
تستخدم شركات التخليق الحمض النووي بشكل روتيني-برامج فحص الأمن في الأمن لضمان عدم إنشاء تسلسلات خطيرة عن غير قصد. لكن ورقة نشرت في علوم في 2 أكتوبر يصف الضعف المحتمل في سير العمل هذا1. إنها تفاصيل كيف يمكن أن تحايل استراتيجيات تصميم البروتين بمساعدة الذكاء الاصطناعي (AI) على برنامج الفحص الذي تستخدمه العديد من شركات تكوين الحمض النووي لضمان تجنب إنتاج التسلسلات عن غير قصد ترميز البروتينات أو مسببات الأمراض الضارة. استخدم الباحثون مقاربة من عالم الأمن السيبراني: "فريق Red Teaming" ، حيث يحاول فريق واحد اختراق دفاعات أخرى (بمعرفتهم).
beiruttime-lb.com
October 3, 2025 at 7:00 AM
صيادون Biothreat يلتقطون الحمض النووي الخطير قبل أن يصنع
تستخدم شركات التخليق الحمض النووي بشكل روتيني-برامج فحص الأمن في الأمن لضمان عدم إنشاء تسلسلات خطيرة عن غير قصد. لكن ورقة نشرت في علوم في 2 أكتوبر يصف الضعف المحتمل في سير العمل هذا1. إنها تفاصيل كيف يمكن أن تحايل استراتيجيات تصميم البروتين…
تستخدم شركات التخليق الحمض النووي بشكل روتيني-برامج فحص الأمن في الأمن لضمان عدم إنشاء تسلسلات خطيرة عن غير قصد. لكن ورقة نشرت في علوم في 2 أكتوبر يصف الضعف المحتمل في سير العمل هذا1. إنها تفاصيل كيف يمكن أن تحايل استراتيجيات تصميم البروتين…
Biothreat hunters catch dangerous DNA before it gets made
DNA-synthesis firms routinely use -biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. But a paper published in Science on 2 October describes a potential vulnerability in this workflow1. It…
DNA-synthesis firms routinely use -biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. But a paper published in Science on 2 October describes a potential vulnerability in this workflow1. It…
Biothreat hunters catch dangerous DNA before it gets made
DNA-synthesis firms routinely use -biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. But a paper published in Science on 2 October describes a potential vulnerability in this workflow1. It details how protein-design strategies aided by artificial intelligence (AI) could circumvent the screening software that many DNA-synthesis firms use to ensure that they avoid unintentionally producing sequences encoding harmful proteins or pathogens. The researchers used an approach from the cybersecurity world: ‘red teaming’, in which one team attempts to break through another’s defences (with their knowledge). They found that some screening tools were unprepared to catch AI-generated protein sequences that recreate the structure, but not the sequence, of known biothreats, says Eric Horvitz, chief scientific officer at Microsoft in Redmond, Washington.
n24usa.com
October 3, 2025 at 5:23 AM
Biothreat hunters catch dangerous DNA before it gets made
DNA-synthesis firms routinely use -biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. But a paper published in Science on 2 October describes a potential vulnerability in this workflow1. It…
DNA-synthesis firms routinely use -biosecurity-screening software to ensure that they don’t inadvertently create dangerous sequences. But a paper published in Science on 2 October describes a potential vulnerability in this workflow1. It…