TL;DR
Get workout gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
A MedPage Today opinion article argues that AI-designed viruses merit safeguards, while naturally emerging viruses currently pose a greater pandemic risk. Its author calls for oversight of AI research and stronger monitoring of viruses in animals, but the piece does not present new evidence that an AI-generated virus has caused an outbreak.
Virologist James Alwine argues that AI-designed viruses deserve safety rules, but that naturally emerging viruses pose the greater current pandemic risk. In an October 14 opinion article for MedPage Today, he points to the laboratory work needed to turn a computer design into a virus and the ongoing risk of animal-to-human transmission; the article is an argument about preparedness, not a report of an AI-created virus causing disease.
Alwine’s argument follows a recent paper in Science that, according to his account, used information about known viruses to design novel bacteriophages, which infect bacteria. He says this capability should be taken seriously and urges scientific societies to establish guidelines and safeguards for AI use in virology and biology. The supplied report does not describe the paper’s methods or establish that the designed viruses pose a danger to people.
He contrasts the design process with viral emergence in nature. A computer-generated design must still be converted into a physical virus through laboratory synthesis, growth and testing, he writes, a process that may yield only a small number of functional viruses. By contrast, zoonotic transmission—the movement of pathogens from animals to people—occurs repeatedly, and most such infections do not spread further. Alwine warns that an infection capable of sustained human-to-human transmission could instead trigger an epidemic or pandemic.
To illustrate the historical consequences, Alwine lists at least 10 animal-derived viral epidemics or pandemics over the past 108 years, including the 1918 influenza pandemic, HIV/AIDS, Ebola and COVID-19. He estimates these events claimed more than 100 million lives, with many more people experiencing lasting harm. These are figures and characterizations presented in the opinion article; it does not provide a separate breakdown or methodology for the total.
The comparison matters for how research and public-health resources are prioritized. AI-designed pathogens raise questions about screening, laboratory controls and responsible research. Zoonotic threats, meanwhile, call for surveillance of animals and people, field research, and systems able to detect unusual infections early. Alwine’s central point is that preparing for one risk should not displace attention to the other.
His account also distinguishes designing a virus from producing and spreading one. A design alone is not an outbreak: it must be made into a viable virus, and then it would need a way to infect and transmit. The article describes these steps as substantial laboratory work, but does not quantify the difficulty or assess how AI capabilities may change it. The practical implication is to build safeguards while continuing to monitor natural viral populations.
Products Worth Considering
Garmin Forerunner 165 GPS Running Smartwatch, 43 mm, Black
Coway Mighty AP-1512HH(W) Air Purifier, White
Garmin Forerunner 55 GPS Running Smartwatch, Black
As an affiliate, we earn on qualifying purchases.
Lessons From Earlier Biology Oversight
Alwine compares the current debate to concerns over recombinant DNA and cloning in the 1970s. He says scientists responded by calling a moratorium and meeting at the 1975 Asilomar Conference, which led to oversight and guidelines for biological research. He calls for a similar pause and conference focused on AI use in virology and biology. This is his policy proposal, not an announcement that a moratorium has been adopted.
The article links zoonotic risk to changes that bring people and animals into closer contact, including habitat destruction, urbanization, shifts in animal behavior and location, wildlife trade, and wet markets. Alwine says scientific evidence supports a wet-market origin for SARS-CoV-2, while acknowledging that the virus’s origin is debated. That attribution remains disputed; the opinion piece does not resolve it.
“Nature is infinitely better at creating viruses than humans and computers are.”
— James Alwine, virologist, in a MedPage Today opinion article
Limits of the AI Risk Assessment
The supplied article does not establish how likely AI-designed viruses are to become viable, infect humans or spread between people. It discusses bacteriophages and does not report an AI-generated virus causing human illness or an outbreak. The relative risk is presented as Alwine’s assessment, not as a quantified comparison of probabilities.
It is also unclear what form any proposed moratorium or safeguards would take, who would enforce them, or whether scientific societies or governments will act. The article calls U.S. monitoring and pandemic preparation inadequate, but does not provide a detailed assessment of existing programs or their current capacity. The origin of SARS-CoV-2 remains contested, as the author notes.
Safeguards and Surveillance Proposals
Alwine urges scientific societies to develop AI-related guidelines and calls for a moratorium and conference on AI in virology and biology. He also argues for field research and ongoing monitoring of viruses in animal populations, alongside stronger pandemic preparation. The article does not give a timeline for those steps or identify a formal process already underway.
For readers, the next development to watch is whether researchers, scientific bodies and public agencies translate these proposals into specific review standards, laboratory safeguards and surveillance plans. Until such measures are announced, the article’s recommendations remain a call to action rather than a new policy or confirmed change in public-health practice.
Key Questions
Did AI create a virus that is causing an outbreak?
No such outbreak is reported in the supplied article. It describes research designing bacteriophages, viruses that infect bacteria, and does not say an AI-generated virus has caused human disease.
Why does Alwine say natural viruses are the greater risk?
He argues that viruses continually mutate and replicate in nature, while a computer design must be made into a viable virus through laboratory work. He also points to repeated animal-to-human spillovers and their potential to lead to sustained transmission.
What safeguards does the author propose?
Alwine calls for scientific societies to create guidelines and safeguards for AI in biology, and proposes a moratorium and conference modeled on the 1975 Asilomar meeting. No such new moratorium is announced in the article.
What remains uncertain about AI-designed viruses?
The article does not quantify the likelihood that an AI-designed virus could be made viable, infect people or spread between them. It also does not establish how AI-related risks compare numerically with zoonotic risks.
Source: rss
