Semaglutide (sold as Ozempic and Wegovy) likely suppresses addiction cravings through a brain structure called the lateral septum, which is dense with GLP-1 receptors. In May 2026, a randomized trial published in The Lancet showed 108 adults with alcohol use disorder and obesity who took semaglutide for 26 weeks cut their heavy drinking days by 41 percentage points, versus 26 for placebo. The VA launched a larger trial on July 28 — enrolling 600+ veterans across 18 medical centers — to test the same drug for AUD. Semaglutide is FDA-approved for Type 2 diabetes and weight loss. It's not approved for addiction treatment.

1. The Trial Data Is Convincing (Anders Fink-Jensen, Nora Volkow, Sarah Carstens)

Addiction medicine has waited decades for a new drug — and semaglutide may be it.

Semaglutide beat placebo by nearly 15 percentage points. Anders Fink-Jensen, D.M.Sc., senior author of the Lancet trial and a researcher at Mental Health Center Copenhagen, says the findings "align with earlier preclinical and clinical research" supporting GLP-1 receptor agonists as a promising treatment approach. Nora Volkow, Director of the National Institute on Drug Abuse and a co-author on the trial, has argued separately that GLP-1 drugs represent a genuinely new class of addiction intervention — not a variation on existing options.

Semaglutide seems to quiet cravings across multiple substances, not just alcohol. A BMJ study of 600,000+ VA patients with type 2 diabetes found that those on GLP-1 drugs — compared to those on a different diabetes drug class — were 18% less likely to develop AUD, with reduced risk for cannabis (14%), cocaine and nicotine (20%), and opioids (25%). University of Otago researcher Robert Munn says the lateral septum acts as the brain's "craving center" and that GLP-1 activation disrupts reward signaling across addictions, not just food. Sarah Carstens, addictions clinical director at Penn Medicine Princeton House Behavioral Health, calls the findings "very positive for the field."

Just 2.6% of Americans with AUD get any medication treatment. The existing options — naltrexone, disulfiram, acamprosate — are chronically underused, and the field has been waiting for something new. Six hundred veterans across 18 sites is a much bigger bet than the Copenhagen team's 108-person pilot. VA Secretary Doug Collins described CRAVE as research the VA is "uniquely situated to launch."

2. But the Evidence Still Has Big Gaps (Christian Hendershot, Mateus d'Ávila, Pat Aussem)

The trial was small, the population was narrow, and two labs can't agree on how the drug works.

The trial only tested people with obesity. All 108 participants had a BMI of 30 or higher. Lancet commentators Christian Hendershot and Klara Klein urged "cautious optimism" and called specifically for dose-response studies and trials in non-obese AUD populations. The trial excluded anyone with AUD who wasn't also overweight.

Two major labs just reached opposite conclusions about how the drug works. A Yale study published August 10 in PNAS found sustained semaglutide treatment activates AgRP hunger neurons — the prevailing assumption was that it suppresses them. Lead researcher Mateus d'Ávila said the finding "completely changes how we think about the mechanism." A Northwestern study from August 2025 in the Journal of Clinical Investigation found the opposite: semaglutide silences those neurons. As of August 2026, neither team has resolved the conflict.

It's too soon to call GLP-1s a solution for addiction. Pat Aussem, LPC, Vice President of Consumer Clinical Content Development at the Partnership to End Addiction, says scientists caution it's "too soon to say if GLP-1 RAs are the solution for SUD." Lucy Connery and Brooke Langella, MPH, at the MATTERS Network add that "it may be years before these medications are sufficiently understood or recognized as appropriate for addiction treatment."

3. And Even Approval Won't Reach Most People Who Need It (Penn LDI Researchers, MATTERS Network, Grewal)

No coverage exists for this use, the drug costs over $1,000 a month at list price, and some courts actively steer patients away from medication.

Right now, no insurer covers semaglutide for alcohol use disorder. Without FDA approval for that indication, insurance denial is the default. Wegovy's list price is $1,349/month; Ozempic runs about $1,000/month. Penn LDI researchers Matthew Klebanoff, M.D. and Jalpa Doshi, Ph.D. have documented that even for weight loss — a condition with FDA approval — coverage is eroding: California's Medi-Cal and Pennsylvania's Medicaid both cut GLP-1 coverage for weight loss in January 2026.

Seven in ten Americans already see GLP-1s as a rich person's drug. A November 2025 survey by The Health Management Academy (n=1,827) found over 70% of Americans believe GLP-1s are accessible only to higher-income people; only 15% said they're accessible to anyone who needs them. People with AUD are disproportionately uninsured and low-income — the population least able to absorb a four-figure monthly drug cost without coverage.

In some courtrooms, choosing semaglutide would count against you. University of Iowa law professor Amandeep S. Grewal documented in a 2026 Health Matrix review that drug court programs routinely pressure participants toward Alcoholics Anonymous and abstinence-based programs, not medication-assisted treatment — and that some laws actively discourage MAT as a condition of avoiding incarceration. Semaglutide would hit the same wall that naltrexone and buprenorphine have long faced. In drug courts, formal evidence matters less than what the judge says.

Where This Lands

Fink-Jensen and Volkow proved the concept: semaglutide plus therapy cuts heavy drinking better than therapy alone. The VA's CRAVE trial will tell us whether that holds at scale and whether non-obese patients benefit. What the trial can't resolve is the contested brain mechanism, the insurance coverage gap, or the courtrooms that still treat medication as cheating. Winning those three fights is how semaglutide actually reaches the 27.9 million Americans with AUD. None of them are pharmacology problems.

Sources