— Nature is far better at creating viruses than computers, but our leaders don't seem to care
by James Alwine, PhD
October 14, 2026 • 3 min read
The news is full of predictions of the doomsday capabilitiesopens in a new tab or window of artificial intelligence (AI). One potential worry is the use of AI to design dangerous viruses. In fact, a recent article in Scienceopens in a new tab or window showed that by using information about known viruses, AI could design novel viruses -- in this case creating bacteriophages, viruses that infect bacteria.
As a virologist, I think we must take this very seriously. Scientific societies should immediately start generating guidelines and safeguards for the use of AI in virology and biology.
Life scientists have always been at the forefront of monitoring and controlling the technologies they use. The emergence of AI-generated viruses is reminiscent of the emergence of recombinant DNA technology and cloning in the 1970s. At that time, concerned scientists called a moratorium and met at the 1975 Asilomar Conferenceopens in a new tab or window, resulting in oversight and guidelines for biological research that are still in force today. We need a similar moratorium and conference for AI's use in virology and biology.
But we also need to put this in perspective. While we should worry about AI-generated viruses, it's useful to compare AI generation of viruses to what happens in nature.
To produce a virus from AI-generated designs, the information must be converted into a real virus. This requires quite a bit of laboratory work to synthesize, grow, and test the viruses, and this may produce only a few functional viruses.
However, every hour of every day, humans encounter new viruses from nature through transfer from animals to humans; this is called zoonosis. While most of these infections are dead ends, any one of them could potentially acquire human-to-human spread and spark an epidemic or pandemic.
In the past 108 years there have been at least 10 animal-derived viral pandemics or epidemicsopens in a new tab or window. They claimed over 100 million livesopens in a new tab or window, not to mention the many millions who were permanently damaged by the infections. These include Spanish Influenza (1918-1920), Asian flu (1957-1958), Hong Kong flu (1968-1970), HIV/AIDS (1981-present), SARS-CoV-1 (2002-2004) H1N1 swine flu (2009-2010), MERS (2012-present), Ebola (2014-present), Zika (2015-present), and SARS-CoV-2 (2019-present). (Although the origin of SARS-CoV-2 is debated, growing scientific evidenceopens in a new tab or window supports its zoonotic origin.)
In nature, viruses have multiple means to randomly mutate, thus enabling them to quickly create millions of variants while infecting and spreading in animals or people. Some of these random changes weaken the variant and it does not survive; other changes make no difference. However, now and then a random change makes a variant that is "better"; for example, it is more efficient at infecting us or at evading our immunity. These more efficient variants are selected and propagate; they become the source of more variants, and the evolution of the virus proceeds.
Nature is infinitely better at creating viruses than humans and computers are. At this moment, the generation and spread of viruses in nature, by random change and worldwide mass replication, remains a far greater threat than AI-designed viruses.
We don't know whether AI design and laboratory technologies can ever catch up to nature and zoonosis. But we do know from prior experience that new viruses from nature can cause death or disability for millions. Yet, our federal leaders are not preparingopens in a new tab or window for a future pandemic and we are not monitoringopens in a new tab or window the emergence of novel viruses in nature. Field researchopens in a new tab or window is essential to understand zoonotic threats. Dynamic viral populations must be monitored and defined in situ. This information is vital to ensure effective pandemic preparation.
We must be aware that viral emergence through zoonoses frequently occurs when host species change their habits and locations due to environmental changes. Such changes result in habitat destruction, urbanization, and increased contact with animals, both wild and domestic. In addition, the international wildlife trade and wet markets provide greatly increased chances of zoonotic infections. Scientific evidence suggestsopens in a new tab or window that a wet market was the origin of SARS-CoV-2.
The U.S. is presently unaware of what is happening with viruses in nature; we are unprepared for the inevitable next zoonotic epidemic or pandemic. The government's denial of science and nonchalance toward both known and emerging infectious diseases leaves us in great danger. Most insidiously, it keeps all too many Americans unaware of nature's awesome power to maintain the constant threat of naturally occurring diseases.