A Virus Scientists Have Watched for Decades
H5N1 avian influenza has circled the globe since the late 1990s, killing hundreds of millions of poultry and occasionally jumping to humans with a terrifying case fatality rate above 50%. For years, the virus struggled to spread between mammals. That changed in 2024-2025 when H5N1 tore through dairy cattle herds across the United States, infected cats, raccoons, and marine mammals, and began showing genetic mutations associated with mammalian adaptation. Public health officials are watching more closely than at any point since the H1N1 pandemic of 2009.
What Makes H5N1 So Dangerous
Influenza viruses mutate rapidly, and H5N1 has two features that worry virologists. First, its lethality: in the roughly 900 confirmed human cases since 2003, over half died. Second, recent isolates from cattle and poultry in 2025 carry mutations in the PB2 gene (particularly E627K and D701N) that improve replication in mammalian cells. These same mutations appeared in the 1918 pandemic virus. The virus has not yet acquired efficient human-to-human airborne transmission, but each new mammalian host gives it another opportunity to pick up the remaining mutations needed. The Science Behind Intimacy and Connection provides context on how viral transmission between species works.
The Dairy Cattle Outbreak No One Expected
Before 2024, nobody thought cows were susceptible to bird flu. When H5N1 appeared in Texas dairy herds in March 2024, it caught veterinary scientists off guard. By early 2026, the virus had been detected in herds across dozens of US states and in several Canadian provinces. Farmworkers exposed to infected cattle developed mild conjunctivitis and respiratory symptoms, and at least one case of severe pneumonia was confirmed in a worker with prolonged, unprotected exposure. The route of infection appears to be primarily through contaminated milk and respiratory droplets from coughing cows.
Vaccine Preparedness
Unlike COVID-19, we are not starting from scratch. H5N1 vaccine candidates have existed for years, and updated versions matched to the current clade (2.3.4.4b) are in production. CSL Seqirus, Sanofi, and GSK all have pre-pandemic stockpile contracts with governments. Canada’s National Advisory Committee on Immunization has drafted contingency plans for mass vaccination of poultry workers, veterinarians, and first responders. mRNA platforms developed during COVID could produce updated vaccines within weeks of a pandemic declaration. The infrastructure built during the coronavirus response is the single biggest advantage we have.
Surveillance Gaps Remain
The biggest concern among epidemiologists is not the virus itself but our ability to detect it early. Can Rubbing Alcohol Kill Lice? Why This Home Remedy Falls Short highlights how surveillance systems are often underfunded. Testing of dairy cattle was initially voluntary in the US, and many farmers resisted testing for fear of economic losses. Wastewater surveillance has proven useful for tracking influenza trends in communities, but rural areas lack the monitoring infrastructure common in cities. Wild bird surveillance, critical for tracking the virus’s geographic spread, depends on a patchwork of academic and government programs with inconsistent funding.
Preparing Without Panicking
Public health messaging around H5N1 walks a tightrope. Overreacting could cause unnecessary fear and economic disruption to the poultry and dairy industries. Underreacting could leave populations unprepared if the virus does achieve efficient human spread. The consensus among infectious disease experts in 2026 is clear: the probability of H5N1 causing a pandemic in any given year is low, but the consequence if it does is catastrophic. That math demands serious, sustained preparation even when the headlines fade.
