The VA Just Bet on Ozempic for Alcoholism — Because No One Ever Built Alcoholism a Drug of Its Own

Today, August 17, 2026, the VA opened recruitment for a 24-week, placebo-controlled trial testing semaglutide against alcohol use disorder across 18 VA medical centers, aiming to enroll more than 600 veterans ages 18 to 80 with moderate-to-severe AUD. More than 400,000 veterans nationally carry that diagnosis, and until this month almost none of them had a medication pipeline built for their disease specifically.

The best new alcoholism treatment in a generation was invented for people trying to lose weight.

That’s not a knock on the science. In a randomized controlled trial published in The Lancet, Copenhagen University Hospital researchers gave 108 people with AUD and comorbid obesity — all also receiving CBT — either weekly semaglutide or placebo. The semaglutide group saw significantly larger drops in heavy drinking days, monthly consumption, drinks per drinking day, and self-reported craving. NIAAA co-investigators worked the study alongside them, and NIAAA Director George Koob told NIH the results could be a “gamechanger” for closing the treatment gap in a country where only a small fraction of the roughly 1 in 9 adults with AUD ever receive an FDA-approved medication for it. Now UW Medicine has joined a 30-site Phase 3 trial of brenipatide, a second-generation GLP-1/GIP drug, for the same indication.

Stack those up and you get something rarer than a promising drug: independent confirmation. A Copenhagen hospital, the federal government’s own addiction research institute, the VA, and a major academic medical center are not reading from the same press release. They arrived at the same drug class from different directions, at the same time, without coordinating.

That convergence is the story. It’s also an indictment.

Naltrexone, acamprosate, and disulfiram are the only medications the FDA has approved for AUD, and the newest of the three is older than the internet. No pharmaceutical company built a serious AUD pipeline in the decades between them and now — not because the biology was unsolvable, but because the payer economics never rewarded it the way obesity and diabetes did. Novo Nordisk and Eli Lilly poured billions into GLP-1 drugs to treat metabolic disease. The alcohol effect showed up as a side note in trial data, noticed almost by accident by patients who mentioned they’d stopped wanting to drink. Addiction medicine didn’t get its own moonshot. It got the spillover from someone else’s.

None of this makes semaglutide an approved AUD treatment yet — that’s precisely what the VA trial and the brenipatide study exist to determine, at population scale, in the population that needs it most. But it’s worth sitting with what it says that the most rigorous, multiply-replicated signal in alcohol treatment in years came from a molecule nobody set out to build for drinkers.

The medicine finally arriving for people with alcohol use disorder wasn’t designed for them. It was rerouted to them — and that’s a very different story than the one that gets told at a ribbon-cutting.