New scientific breakthroughs make it increasingly easy to identify dangerous viruses in nature, manipulate them in the lab and synthetically create them from genetic sequences. But some scientists have taken it further, adding "gain of function" mutations that make potential pandemic viruses more transmissible. The National Institutes of Health funded two research groups to increase the transmissibility of an earlier strain of avian influenza that had killed hundreds of people but could not efficiently spread from person to person. Both groups created viral mutants that could transmit in ferrets. The Obama administration was so alarmed that it halted gain-of-function work on potential pandemic influenza viruses in 2014, but the N.I.H. allowed it to restart by 2019. In my view, there is no justification for intentionally making potential pandemic viruses more transmissible. The consequences of an accident could be too horrific, and such engineered viruses are not needed for vaccines anyway. Natural viruses that haven't yet infected humans can also pose a risk if researchers try to find the most dangerous ones and bring them back to the lab for experiments. Suspicions about a lab-accident origin of SARS-CoV-2 have been fueled by the fact that the Wuhan Institute of Virology was involved in Chinese and international efforts to find and experiment with new high-risk coronaviruses. A final category of pandemic risk involves viruses that used to transmit in humans but became extinct long ago " like the 1918 influenza virus.