When I started suing food companies more than thirty years ago, after E. coli O157:H7 in undercooked Jack in the Box hamburgers killed four children and sickened hundreds, proving where a pathogen came from was slow, painstaking, and often impossible. Investigators matched bacteria using crude “fingerprinting” methods that could tell you two samples looked similar, but rarely that they were truly the same organism.
That world is gone. Whole genome sequencing — reading the complete DNA of a bacterium — has done for foodborne illness what DNA testing did for criminal forensics. Today, when Listeria turns up in a nursing-home resident in California and in a frozen nutritional shake produced two thousand miles away, public-health scientists can say with near-certainty it is the same organism. The FDA’s GenomeTrakr and the CDC’s PulseNet databases and NCBI’s website now hold hundreds of thousands of these genomic fingerprints. The 2025 Lyons Magnus outbreak — 42 sickened, 14 dead across 21 states — was solved this way. So was the largest and deadliest listeriosis outbreak in history: the 2017–2018 South African catastrophe that killed 216 people and was traced, through a single sequence type called ST6, back to polony from one Tiger Brands plant in Polokwane.
For public health, the case is overwhelming. WGS finds the source faster, thereby making outbreaks smaller, narrows recalls so they strike the guilty product instead of an entire commodity, and catches contamination at vanishingly low levels. None of that is seriously in dispute.
The harder question — and the one companies actually ask me — is whether a food producer should run that same technology on its own products and inside its own plant.
Here is the uncomfortable part, and I’ll say it plainly because the industry’s lawyers usually won’t: a positive result you generate today can be used against you tomorrow. If you swab your plant and find Listeria in a floor drain, sequence it, and that genome later matches a sick child’s, you are tied to that case in a way no jury will forget. Some defense attorneys take the logical next step and quietly advise their clients to test less — because what you don’t know can’t be subpoenaed.
I have spent decades on the other side of that advice, and I am here to tell you the bad advice is both morally bankrupt and, in the long run, a legal disaster.
Choosing not to look is not a defense. It is willful blindness, and juries understand it instantly. The company that finds a resident strain, documents it, and surgically eliminates it has a contemporaneous record of doing exactly what a responsible food producer is supposed to do. The company that decided ignorance was safer has a record of nothing — until the outbreak arrives anyway, larger and deadlier for having gone undetected, and a plaintiff’s lawyer like me asks under oath why a company with hundreds of millions in revenue chose not to use a technology that a public-health lab in a developing nation could afford.
And WGS cuts both ways. The same precision that can implicate you can clear you. When your product is swept into an outbreak it had nothing to do with, your genomic data is the fastest route to proving your strain is not the strain — to getting your name out of the headline and your trucks back on the road. I have watched sequencing exonerate an accused producer as decisively as it has condemned a guilty one.
WGS is not magic. It demands real bioinformatics expertise, validated pipelines, and reference databases that still have gaps for rare and emerging pathogens.
But none of that is an argument for staying in the dark. The food industry’s reflex has been to know as little as possible about its own contamination, on the theory that what it didn’t know couldn’t hurt it in court. The genome has ended that bargain. The pathogens in your plant already have a fingerprint. The only question is whether you are willing to read it before someone’s child does.
