Here’s a sentence that would have gotten you laughed out of a conference five years ago: we just put CRISPR inside a virus, fed it to humans, and it killed exactly the bacteria we told it to — and nothing else.

That’s not a thought experiment. That’s the Phase 1 data SNIPR Biome just published in The Lancet Microbe.

SNIPR001 is an oral cocktail of four CRISPR-engineered bacteriophages designed to find E. coli in the gastrointestinal tract, deliver a CRISPR-Cas payload that shreds the bacterial DNA, and leave everything else — every beneficial microbe, every commensal species — completely untouched. In a randomized, placebo-controlled, first-in-human study across three dose levels, SNIPR001 was safe, well tolerated, showed dose-proportional recovery in stool, zero meaningful systemic exposure, and — the kicker — stable overall gut microbiome composition.

Read that last part again. You can take a CRISPR-armed phage, swallow it, and it kills the bad bacteria without nuking your microbiome. The antibiotic era’s original sin — collateral damage to the gut — just got a potential answer.

SNIPR Biome is Danish, Copenhagen-based, and has been quietly building toward this moment since 2017. They hold over 60 granted patents worldwide, have raised $139 million across nine rounds, and were the first company to orally dose humans with a CRISPR therapeutic. Let that sink in — every other CRISPR company is injecting, infusing, or implanting. SNIPR handed it to patients in a capsule.

The lead indication is practical and urgent: preventing bloodstream infections in hematological cancer patients undergoing stem-cell transplants who are colonized with fluoroquinolone-resistant E. coli. These patients are immunocompromised, gut bacteria can translocate into the blood, and antibiotic-resistant strains make standard prophylaxis unreliable. A Phase 1b trial in exactly this population is already more than halfway enrolled, co-funded by CARB-X.

But the platform is the real story. SNIPR’s CRISPR-Guided Vectors can be programmed against different bacterial targets — they’re already developing a Pseudomonas aeruginosa program for cystic fibrosis, and the Bill & Melinda Gates Foundation funded a microbial gene therapy program for environmental enteric dysfunction in pregnant women. The same platform that kills bacteria can also deliver therapeutic genes to the microbiome. That’s not an antimicrobial. That’s a programmable biological platform for the gut.

Meanwhile, Locus Biosciences is pushing AI-designed phage therapy into hospital-acquired pneumonia. The phage therapy field is heating up — but SNIPR is the only one with published human CRISPR-phage data in a top-tier journal. That matters when you’re trying to convince regulators, investors, and clinicians that this isn’t science fiction anymore. It’s in The Lancet.