NIH Montana Lab Accidents Expose Protocol Gaps for Lethal Viruses

Monkey With Deadly Virus Bites Worker at US Government Lab — Sabine Hossenfelder

Sabine Hossenfelder

While public anxiety centers on AI-engineered bioweapons, sheer human incompetence and absent emergency protocols inside government containment facilities pose the most immediate biological threat.

AI-assisted summary of Monkey With Deadly Virus Bites Worker at US Government Lab — original source. Check the original for full context.

Source published:

Key ideas

  • A laboratory monkey infected with Crimean-Congo hemorrhagic fever bit through a technician's protective gear at an NIH Biosafety Level 4 facility in Montana, revealing that administrators lacked standard operating protocols to manage the emergency. (3:00)
  • Following the exposure incident, the chief of the same Montana lab's virus ecology section was caught attempting to smuggle over 100 biological vials containing deactivated Mpox specimens through US customs. (4:31)
  • Sabine Hossenfelder argues that while AI-assisted bio-engineered pathogens pose humanity's highest extinction risk, everyday human incompetence in institutional research labs represents the far more immediate hazard. (1:58)
  • Partisan gridlock over federal science funding—ranging from conservative calls to abolish the NIH to defenders who dismiss administrative failures—leaves critical laboratory biosafety vulnerabilities unaddressed. (4:31)

Detailed notes

## Background Published on YouTube by Sabine Hossenfelder on August 7, 2026, this video investigates serious safety breaches at a United States government laboratory. The video description highlights an incident where a lab worker was bitten by a monkey carrying a deadly virus without any protocols in place to handle it, alongside an attempt by lab personnel to smuggle outbreak samples into the country while public attention remains fixated on AI bioweapons.

## TL;DR Sabine examines the stark contrast between public panic over AI-engineered bioweapons and the reality of basic human incompetence within high-containment biosafety facilities. She details a November 2025 incident at an NIH-operated BSL-4 facility in Montana, where a worker bitten by a monkey infected with Crimean-Congo hemorrhagic fever faced severe bureaucratic delays in medical transfer and diagnostic sample testing due to an absence of standard operating protocols. Compounding these failures, the laboratory's virus ecology section chief was subsequently caught attempting to smuggle over 100 biological vials—including deactivated Mpox specimens—through customs. Contextualizing these events with historical laboratory leaks from the UK, China, and the Soviet Union, Sabine cautions that neither partisan denial of institutional dysfunction nor calls to abolish research bodies solve the underlying management crisis.

## Atmosphere & style The presentation blends Sabine's signature sardonic, matter-of-fact skepticism with a chilling, urgent examination of catastrophic biological risks. The pacing transitions smoothly from a relaxed, personal sponsored segment into an alarming, forensic breakdown of administrative paralysis and logistical farce. It feels soberingly pragmatic and unsettling, exposing how fragile global biosecurity truly is when undermined by mundane bureaucratic blunders.

## Bulletin - [03:00](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=180s) A laboratory monkey infected with Crimean-Congo hemorrhagic fever bit through a technician's protective gear at an NIH Biosafety Level 4 facility in Montana, revealing that administrators lacked standard operating protocols to manage the emergency. - [04:31](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=271s) Following the exposure incident, the chief of the same Montana lab's virus ecology section was caught attempting to smuggle over 100 biological vials containing deactivated Mpox specimens through US customs. - [01:58](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=118s) Sabine Hossenfelder argues that while AI-assisted bio-engineered pathogens pose humanity's highest extinction risk, everyday human incompetence in institutional research labs represents the far more immediate hazard. - [04:31](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=271s) Partisan gridlock over federal science funding—ranging from conservative calls to abolish the NIH to defenders who dismiss administrative failures—leaves critical laboratory biosafety vulnerabilities unaddressed.

## Key takeaways - [01:58](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=118s) A bio-engineered pathogen that combines high infectivity with lethality currently represents the highest existential risk to human civilization. - [01:58](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=118s) The tools required to assemble lethal biological agents are widely accessible, and artificial intelligence is rapidly lowering the technical barrier to utilizing them. - [01:58](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=118s) Frontier AI models like Claude 3.5 Sonnet feature intense guardrails on biology queries, yet public concern over AI bioweapons often ignores the more urgent threat of routine human error. - [03:00](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=180s) An NIH BSL-4 facility worker in Montana was bitten by a monkey infected with Crimean-Congo hemorrhagic fever, which carries a human fatality rate of 10% to 40%. - [03:00](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=180s) A separate exposure risk occurred at the same facility just three months later when a second technician changed cages housing infected mice. - [03:00](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=180s) Emergency response to the monkey bite devolved into administrative chaos, requiring ad-hoc travel approvals from multiple state governors to move the patient by ambulance. - [03:00](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=180s) Attempts to send patient diagnostic blood samples out of state were stalled because officials struggled to find an authorized airplane to fly the cargo. - [03:00](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=180s) Internal whistleblower reports revealed that the NIH facility still lacked standardized emergency protocols seven months after the bite occurred. - [04:31](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=271s) The head of the facility's virus ecology section misled customs agents by claiming a transport case contained only diagnostic hardware, despite holding over 100 Mpox vials. - [04:31](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=271s) Political polarization impedes lab oversight: critics demand deep budget cuts or abolition of the NIH, while defenders refuse to acknowledge mismanagement and waste. - [04:31](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=271s) Historical precedents—including anthrax in the USSR, smallpox in Birmingham, SARS in China, and foot-and-mouth disease in the UK—demonstrate that institutional lab leaks are recurring real-world events. - [04:31](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=271s) Dismantling scientific agencies is counterproductive because society cannot afford the resulting research vacuum, yet institutional mismanagement must be actively reformed across both the US and Europe.

## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=0s) Introduction and Personal DNA Discovery

A worker inside a United States government high-containment facility was recently bitten by a monkey carrying a deadly virus, exposing the fact that the administration had no routine protocols ready to manage such an emergency. Around the same time, another individual from that exact same lab was caught attempting to smuggle biological outbreak samples into the country. These incidents surface at a fraught cultural moment: concerns over AI-engineered bioweapons are peaking, while political leaders in the US push to slash budgets for the National Institutes of Health (NIH).

Before diving into the lab failures, Sabine shares her experience testing her genetics with video sponsor MyHeritage for €19. The cheek-swab kit revealed that while her ancestry is predominantly Eastern European, it unexpectedly includes 2.8% English lineage along with Dutch roots she was unaware of. Because her parents separated when she was young and her father passed away shortly thereafter, she had had virtually no connection with her paternal relatives. To her surprise, she discovered that an aunt had built out a family tree on the platform, uncovering ancient lineage tracing even further east. She notes that the service also provides trait reports on characteristics like introversion, creativity, and caffeine sensitivity, while maintaining strict privacy policies that prohibit selling or licensing genetic profiles.

[01:58](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=118s) The Existential Threat of Engineered Pathogens vs. Human Incompetence

Looking at catastrophic threats facing humanity, the greatest danger does not necessarily stem from conventional conflict. As Sabine warns, “A bio-engineered pathogen that's both highly infectious and lethal is currently the highest risk for human extinction.” A pathogen of that caliber could realistically decimate most of global civilization, and modern responses to the COVID-19 pandemic demonstrate that institutions simply cannot move fast enough to neutralize a fast-moving biological threat. The foundational tools to create such organisms are already accessible and functionally impossible to police. The primary constraint has long been the technical know-how required for assembly, but artificial intelligence is rapidly dissolving that barrier.

This tension is apparent in frontier AI development. Anthropic’s model, Claude 3.5 Sonnet (referred to in development as "Fable"), was rolled out with guardrails so strict that the system refused to answer routine, benign biological inquiries. Yet hyper-cautious safeguards merely underscore the core dilemma: the foundational knowledge is publicly available, requiring only sufficient processing power and intelligence to assemble it. However, focusing exclusively on rogue actors or AI-assisted bio-terror overlooks the far more pervasive and immediate hazard: sheer human incompetence inside institutional laboratories.

[03:00](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=180s) Chaos at the Montana BSL-4 Facility

On November 13, 2025, an accident occurred inside an NIH-operated Biosafety Level 4 (BSL-4) facility in Montana. A laboratory monkey infected with Crimean-Congo hemorrhagic fever—a severe tick-borne pathogen carrying a human case fatality rate of 10% to 40% according to the World Health Organization—bit through a laboratory worker's personal protective gear. Just three months later, a second technician was potentially exposed to the same lethal virus while swapping out cages housing laboratory mice.

```mermaid flowchart TD accTitle: Montana BSL-4 Exposure Response Breakdown accDescr: Flowchart illustrating the chain of logistical failures after a worker was bitten by an infected monkey.

subgraph Incident["Laboratory Incident"] A["Worker bitten through PPE by infected monkey"] B["Sent to local Montana hospital"] end

subgraph LogisticsFailures["Bureaucratic & Logistical Hurdles"] C["Attempted ambulance transfer to Washington"] D["Delayed: Required approval from multiple state governors"] E["Attempted blood specimen shipment"] F["Delayed: Difficulty securing an air transport plane"] end

subgraph Outcome["Aftermath"] G["Worker miraculously avoided infection"] H["Whistleblower reveals lack of established protocols 7 months later"] end

A --> B B --> C C --> D B --> E E --> F D --> G F --> G G --> H ```

Although the exposed employees ultimately tested negative and escaped infection, internal NIH records leaked by a whistleblower revealed an alarming lack of institutional preparation. Rather than executing a pre-planned emergency procedure, the administration improvised at every turn:

- The bitten employee was initially brought to a local Montana hospital before administrators decided a transfer to specialized care in Washington state was required. - Arranging that ground ambulance transfer ground to a halt because staff had to obtain extraordinary, ad-hoc travel permissions from multiple state governors. - When the facility attempted to fly urgent patient blood samples to an outside diagnostic lab, officials struggled simply to track down an available airplane willing and authorized to transport the cargo. - Seven months following the monkey bite, the whistleblower confirmed that the NIH facility had still failed to formalize a standard operating protocol for similar future exposures, even as leadership publicly maintained that established guidelines had been observed.

[04:31](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=271s) Smuggled Mpox Vials and the Lab Leak Reality

Institutional carelessness escalated further when the chief of the Montana laboratory's virus ecology section returned from a field assignment in the Republic of the Congo. Arriving with research samples collected during an active Mpox outbreak, the scientist and an accompanying colleague claimed to customs agents that their travel case held nothing more than diagnostic hardware. Federal customs inspectors intervened, inspected the container, and uncovered over 100 biological vials, a portion of which held deactivated Mpox virus specimens.

These scandals have broken amidst fierce US political debates over whether to defund or overhaul scientific institutions. Conservatives and public health critics, such as Scott Atlas, have advocated for deep budget cuts or even abolishing the NIH entirely. Conversely, political defenders frequently refuse to acknowledge systemic administrative dysfunction and financial waste within federal agencies, dismissing all conservative critiques as purely anti-science. While both sides remain locked in partisan posturing, institutional vulnerabilities persist unaddressed.

Lab escapes are historical reality rather than fringe speculation: - In the 1970s, an accidental aerosol release of anthrax spores from a Soviet military biology facility killed at least 66 people. - In 1978, a smallpox leak from a medical school laboratory in Birmingham, UK, infected two individuals and killed one. - In 2004, the SARS coronavirus escaped from a Chinese containment laboratory, resulting in multiple infections and one fatality. - In 2007, an environmental leak of foot-and-mouth disease virus from an animal health research center in Pirbright, UK, reached livestock on nearby commercial farms.

Sabine explains that while the impulse to dismantle dysfunctional organizations and rebuild them from the ground up is understandable, society cannot afford the research vacuum that would follow. Furthermore, these management problems are not uniquely American; European institutions suffer from identical oversights, though they receive far less open discussion. Summarizing the stakes facing biotechnology and safety culture, Sabine emphasizes: “When it comes to AI, as much as when it comes to the missing safety procedures in labs, refusing to address problems doesn't prevent them from happening.”

## Notable quotes > [01:58](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=118s) A bio-engineered pathogen that's both highly infectious and lethal is currently the highest risk for human extinction. — Sabine Hossenfelder

> [04:31](https://www.youtube.com/watch?v=3KO9niSQtlQ&t=271s) When it comes to AI, as much as when it comes to the missing safety procedures in labs, refusing to address problems doesn't prevent them from happening. — Sabine Hossenfelder

## Mentioned MyHeritage, National Institutes of Health, Anthropic, Claude 3.5 Sonnet, World Health Organization, Scott Atlas, Crimean-Congo hemorrhagic fever, Mpox, SARS coronavirus

## Video description MyHeritage is running a special promotion right now! Click👉 https://bit.ly/SabineHossenfelder Use the Coupon Code: SABINE to get your DNA kit for just $19!

A worker at a US government lab got bitten by a monkey carrying a deadly virus, and it turned out nobody knew what to do about it. Around the same time, someone from the same lab tried to smuggle outbreak samples into the country. All of this while everyone's worried about AI making bioweapons. Let's take a look.

#science #sciencenews #virus #physics #health

👕T-shirts, mugs, posters and more: ➜ https://sabines-store.dashery.com/ 💌 Support me on Donorbox ➜ https://donorbox.org/swtg 👉 Transcript with links to references on Patreon ➜ https://www.patreon.com/Sabine 📝 Transcripts and written news on Substack ➜ https://sciencewtg.substack.com/ 📩 Free weekly science newsletter ➜ https://sabinehossenfelder.com/newsletter/ 👂 Audio only podcast ➜ https://open.spotify.com/show/0MkNfXlKnMPEUMEeKQYmYC 🔗 Join this channel to get access to perks ➜ https://www.youtube.com/channel/UC1yNl2E66ZzKApQdRuTQ4tw/join 📚 Buy my book ➜ https://amzn.to/3HSAWJW

[Source](https://www.youtube.com/watch?v=3KO9niSQtlQ)

中文简报

NIH蒙大拿实验室事故暴露致命病毒处置规程缺陷 尽管公众的焦虑聚焦于AI设计的生物武器，但政府防护设施内纯粹的人为失职与应急规程的缺失才是最直接的生物威胁。 Sabine探讨了公众对AI设计生物武器的恐慌与高级别生物安全防护设施内基础人为失职现实之间的鲜明对比。她详细介绍了2025年11月发生在美国国立卫生研究院（NIH）位于蒙大拿州的BSL-4实验室的一起事故：一名工作人员被感染了克里米亚-刚果出血热的猴子咬伤，但由于缺乏标准操作规程，在医疗转运和诊断样本检测方面遭遇了严重的官僚主义延误。令这些失误雪上加霜的是，该实验室病毒生态部门主管随后因企图通过海关走私100多支生物试管（其中包括灭活的猴痘标本）而被捕。结合英国、中国和苏联以往的实验室泄漏历史背景，Sabine告诫称，无论是出于党派偏见否认体制功能失调，还是呼吁废除研究机构，都无法解决背后的管理危机。 在NIH位于蒙大拿州的生物安全四级（BSL-4）设施中，一只感染了克里米亚-刚果出血热的实验猴咬穿了一名技术人员的防护装备，暴露出管理层缺乏应对此类紧急情况的标准操作规程。 (3:00) 在该暴露事件发生后，同一蒙大拿实验室的病毒生态部门主管因企图通过美国海关走私100多支装有灭活猴痘标本的生物试管而被抓获。 (4:31) Sabine Hossenfelder认为，尽管AI辅助的生物工程病原体构成了人类最高的灭绝风险，但科研机构实验室中日常的人为失职才是更为紧迫的现实危险。 (1:58) 围绕联邦科研经费的党派僵局——从保守派呼吁废除NIH，到维护者对管理疏漏不以为意——使得关键的实验室生物安全漏洞始终未能得到解决。 (4:31)

Reading summary

Sabine Hossenfelder argues that while artificial intelligence accelerates the catastrophic risks of engineered pathogens, immediate biological danger stems from systemic human incompetence and institutional mismanagement within high-containment laboratories. Catastrophic Biological Risks and AI Capabilities (1:58) A bio-engineered pathogen that combines high infectivity with high lethality represents the greatest existential threat to human civilization, according to Sabine. The global response to COVID-19 showed that institutional mechanisms operate too slowly to halt a rapidly spreading contagion. While the foundational scientific knowledge and tools for synthesizing dangerous pathogens have long been accessible, the technical assembly barrier is now eroding as artificial intelligence advances. Institutional Failures at a Montana BSL-4 Lab (3:00) In November 2025, an animal technician at an NIH-operated Biosafety Level 4 facility in Montana was bitten through protective gear by a monkey infected with Crimean-Congo hemorrhagic fever, a virus carrying a 10% to 40% case fatality rate according to the World Health Organization. Just three months later, another worker faced potential exposure while changing mouse cages. Although neither worker contracted the pathogen, leaked internal records revealed that facility leadership operated without established emergency protocols. Logistical Improvisation and Biosecurity Violations (3:00) The response to the Montana exposure involved severe logistical breakdowns, including delays transferring the patient by ambulance to Washington state due to ad-hoc gubernatorial approval requirements, as well as difficulties finding an aircraft authorized to fly diagnostic blood samples. Internal documents obtained by a whistleblower showed that standard protocols remained unformulated seven months after the incident, despite public statements to the contrary. Smuggling Incidents and Historical Lab Leaks (4:31) Lax institutional practices were further highlighted when the Montana lab's virus ecology chief and a colleague were caught at customs attempting to enter the country with over 100 undeclared biological vials from an Mpox outbreak in the Republic of the Congo, having falsely declared only diagnostic equipment. Sabine underscores that accidental laboratory escapes have precedent: an accidental anthrax spore release in the Soviet Union during the 1970s killed at least 66 people, a 1978 smallpox leak in Birmingham resulted in one death, a 2004 SARS laboratory leak in China caused multiple infections and one fatality, and a 2007 foot-and-mouth disease leak in the UK infected nearby livestock. Sabine notes that while critics like Scott Atlas advocate defunding or dismantling agencies like the NIH, outright abolition creates dangerous research vacuums. Because institutional vulnerabilities and safety oversights also affect European laboratories, she concludes that ignoring administrative dysfunction does not prevent catastrophic breaches from occurring.

精炼阅读摘要

Sabine Hossenfelder 指出，尽管公众对人工智能辅助合成生物武器的忧虑与日俱增，但当前人类面临的最紧迫生物安全隐患实为科研机构内部的人为失职与管理失能；近期美国政府高等级生物实验室发生的一系列事故与应对混乱，暴露出实验室应急机制和安全监管存在系统性漏洞。 人工智能威胁掩盖下的人为管理失能 (1:58) 在评估人类面临的毁灭性威胁时，Hossenfelder 认为，兼具高传染性与高致死率的工程改造病原体是目前导致人类灭绝的最大潜在风险。从全球应对新冠疫情的实际表现来看，现有公共机构的反应速度难以遏制高传播速度的生物威胁。目前构建此类微生物的基础科学工具已广泛普及且客观上难以封禁，过去的核心阻碍在于组装所需的专业技术门槛，而人工智能的快速发展正在迅速消除该壁垒。例如 Anthropic 旗下的 Claude 3.5 Sonnet 模型（开发代号“Fable”）由于设置了过于严苛的安全防护，一度甚至拒绝回答常规的基础生物学咨询。 然而，Hossenfelder 强调，公众和政策制定者若将目光完全锁定在恶意外部行为者或 AI 辅助的生物恐怖主义上，就会掩盖正规科研机构内部更为普遍、更易引发灾难的现实威胁：即高等级生物实验室内严重的制度性懈怠与人为疏漏。 蒙大拿BSL-4实验室暴露与应急崩溃 (3:00) 2025 年 11 月 13 日，美国国立卫生研究院（NIH）在蒙大拿州运营的一处生物安全四级（BSL-4）实验室发生险情：一只感染了克里米亚-刚果出血热（世界卫生组织统计该病病死率为 10% 至 40%）的实验猴咬破了工作人员的个人防护装备；仅三个月后，另一名技术员在更换小鼠笼舍时，也发生了针对同一种致命病毒的潜在暴露事故。虽然两名涉事雇员后续检测均为阴性并侥幸脱险，但举报人披露的内部记录揭示出该机构在事故应急处理上处于完全失控和即兴应付的状态。 记录显示，该设施在事故发生后并无既定应急预案：被咬伤的员工最初被送往蒙大拿当地医院，随后管理层决定通过救护车将其转运至华盛顿州的专业医疗机构，但这趟地面转运因必须临时向沿途多个州的州长逐一申请通行许可而陷入停滞；与此同时，当机构试图将关键血液样本空运至外部机构进行诊断时，官员甚至难以联络到愿意且有权承运的飞机。更具争议的是，举报人证实，在猴咬事件发生整整七个月后，该实验室仍未建立起标准操作流程，而管理层在公开场合却声称所有应急操作均严格遵循了既定规范。 违规携带病毒样本与实验室泄漏历史先例 (4:31) 蒙大拿实验室的违规操作进一步蔓延至涉外样本运输。该实验室病毒生态学部门主管在结束刚果共和国猴痘爆发前线的科研任务回国时，携带了大量实地采集的病毒样本。该主管及同行人员向海关官员谎称其行李箱内仅有诊断硬件，但在联邦海关的开箱检查下，被当场搜出 100 多个生物试管，其中部分装有灭活的猴痘病毒标本。 Hossenfelder 指出，实验室病原体泄漏绝非无根据的猜想，而是有据可查的历史事实：20 世纪 70 年代，苏联一处军事生物实验室意外发生炭疽芽孢气溶胶泄漏，导致至少 66 人死亡；1978 年，英国伯明翰大学医学实验室发生天花病毒泄漏，导致两人感染、一人死亡；2004 年，中国某高等级实验室的 SARS 冠状病毒发生外泄，引发多人感染并造成一人死亡；2007 年，英国皮尔布赖特动物卫生研究所泄漏的口蹄疫病毒则直接扩散至周边商业农场的牲畜群中。 政治两极化争议与科研监管困境 (4:31) 这些实验室违规事件的曝光正值美国就科研机构经费与职能展开激烈政治辩论之际。包括斯科特·阿特拉斯（Scott Atlas）在内的保守派公共卫生批评人士主张大幅削减甚至彻底废除 NIH 的预算与编制；与此同时，联邦机构的坚定拥护者则往往拒绝正视机构内部客观存在的行政失能与经费浪费，将所有外部批评一概驳斥为反科学言论。这种政治两极化导致真正的实验室安全漏洞长期被各方忽视。 Hossenfelder 解释称，全面取缔失职机构的激进主张固然源于对管理混乱的不满，但社会同样无法承受研究完全停摆所带来的巨大科研真空。此外，这类管理漏洞不仅局限于美国，欧洲研究机构内部也存在类似疏漏，却甚少在公共领域展开透明探讨。在生物技术不断发展的同时，回避失误和逃避管理问责并不能阻止下一次潜在泄漏事故的发生。 Hossenfelder 总结认为，面对 AI 降低生物攻击门槛的宏观趋势与实验室内缺乏规范应急机制的现实矛盾，各方不能因政治阵营对立而选择性失明；彻底废止科研机构并非现实选项，但直面制度失灵、建立刚性透明的安全操作规范是防范未来灾难性病原体泄漏的唯一出路。

Source quotes

[1:58] Sabine Hossenfelder: A bio-engineered pathogen that's both highly infectious and lethal is currently the highest risk for human extinction. [4:31] Sabine Hossenfelder: When it comes to AI, as much as when it comes to the missing safety procedures in labs, refusing to address problems doesn't prevent them from happening.

Original transcript

[0:00] A worker inside a US government high-containment  lab recently got bitten by a monkey carrying a [0:07] deadly virus. The lab admin then didn't know  what to do simply because it's got no routine [0:14] procedure in place. Someone else from the same lab  then tried to smuggle biological samples from an [0:21] Mpox outbreak into the US. That comes at a time  when worries about AI-generated bioweapons are at [0:28] an all-time high and the US government calls  to cut funding for the National Institute of [0:34] Health. Let's have a look. I recently took a DNA  test from today's sponsor, My Heritage, and well, [0:41] I didn't expect this. At My Heritage, you get a  DNA analysis for a super fair price of €19. And [0:48] that tells you so much about your origins  and your genetic traits. It's easy to do. [0:53] You get the test kit sent to your home. Just  swap a cotton swab in your mouth and send it [0:58] back to the lab. My results told me that my  genetic origins are mostly from Eastern Europe, [1:04] but include 2.8% English and some Dutch, which I  didn't know anything about. My parents separated [1:11] early and my father died soon after. I have  basically no contact with my father's side of [1:17] the family. But surprise, it turns out that one of  my aunts has set up a family tree on my heritage. [1:25] I also learned that my ancient origins partly  come from even further east. And my heritage also [1:31] offers a lot of trade reports such as creativity,  introversion, and caffeine sensitivity. They [1:37] have a serious data privacy policy and will never  sell or license your genetic data. And of course, [1:43] I have a special offer. If you use my code, you  get a discount on the DNA kit with free shipping, [1:49] plus a 30-day free trial for just €19. Links in  the info below. A bio-engineered pathogen that's [1:58] both highly infectious and lethal is currently the  highest risk for human extinction. Such a pathogen [2:05] could plausibly wipe out most of the human  population. And seeing what happened with COVID, it's [2:11] unlikely we'd be able to counteract the threat  rapidly enough. The technology to engineer such a [2:18] pathogen exists and it's basically impossible to  regulate. The remaining bottleneck is knowledge [2:24] about how to do it and the facilitator to get  this knowledge is artificial intelligence. [2:30] Anthropic's newest AI model, Fable, was equipped  with safety measures so restrictive that it simply [2:37] refused to answer even basic biological questions.  But really, this just highlights the problem, [2:44] which is the knowledge is out there, and it's  just a matter of intelligence to put it together. [2:50] Then again, the risk of malicious actors shouldn't  distract us from the more immediate problem, [2:57] which are simply incompetent actors. Consider  this. On November 13th, 2025, a monkey infected [3:05] with Crimean-Congo haemorrhagic favor bit through  a workers's protective equipment in a Biosafety [3:12] Level Four laboratory in Montana. The World Health  Organization puts the fatality rate of the virus [3:18] between 10 and 40%. 3 months later, another  worker was potentially exposed to the virus [3:24] while changing cages containing laboratory mice.  These accidents happened in a laboratory run by [3:32] the United States National Institutes of Health,  NIH for short. The workers eventually did not [3:39] develop an infection. But the problem is that the  lab administration clearly didn't know what to do, [3:46] which now became clear from an NIH document  provided by a whistleblower who simply said [3:52] they don't have a real protocol. The worker who  was bitten by the monkey was first sent to a [3:58] hospital in Montana. Admin then tried to arrange  an ambulance transfer to a different hospital in [4:05] Washington, but first had to get extra approval  from multiple governors. They tried to send blood [4:11] samples to a laboratory, but had difficulty  finding a plane to get the job done. 7 months [4:16] later, the whistleblower noticed they still hadn't  established the necessary safety protocol, though [4:24] the NIH insists that established procedures were  followed. Meanwhile, the chief of the laboratory's [4:31] virus ecology section returned from the Republic  of Congo with tissue samples related to the Mpox [4:39] outbreak there. According to a criminal complaint,  he and a collaborator told customs officers that [4:45] a case contained diagnostic equipment. Alas,   officials opened it and found more than 100 vials, [4:51] some of which contained deactivated Mpox virus.  This comes at a time when the US government [4:58] has called again to cut spending to the NIH  dramatically on the public health advisor Scott [5:05] Atlas has called to abolish it. Cue the political  left which doesn't want to admit that the NIH has [5:11] serious administrational problems and much of  the money goes to waste and instead insists [5:16] that everything which comes from the political  right is anti-science. And while they argue, [5:22] [ __ ] happens and continues to happen. That  dangerous viruses leak from laboratories is [5:28] not a conspiracy theory. It's a very real problem.  In the 1970s, Anthrax spores escaped from a Soviet [5:36] facility, killing at least 66 people. And Smallpox  viruses escaped from a laboratory in the UK, [5:42] killing one person and infecting a second.  In 2004, SARS escaped from a Chinese lab, [5:49] killing one and causing multiple other infections.  In 2007, a foot and mouth disease virus escaped [5:55] from a lap in the UK and reached farms. Now,  look, I'm very sympathetic to the idea of [6:02] tearing everything down and starting over from  scratch, but this only makes sense if you can [6:07] afford a period in which nothing works. And this  just isn't the case for much of research in the [6:14] NIH as much as elsewhere because make no mistake,  we have the same problems here in Europe. We just [6:21] don't talk about it. And really, I think this  is the main lesson here. When it comes to AI, [6:27] as much as when it comes to the missing safety  procedures in labs, refusing to address problems [6:34] doesn't prevent them from happening. And don't  forget to get your DNA test at My Heritage. Use [6:40] my link and code for a special offer. Links  below. Thanks for watching. See you tomorrow.