Is this weed breathalyzer for real? Don’t hold your breath.

This post is part of our High-tech High series, which explores weed innovations, and our cultural relationship with cannabis, as legalization in several U.S. states, Canada, and Uruguay moves the market further out of the shadows.

Will you soon be blowing into a tube connected to a handheld gadget that can tell an officer or an employer whether you’ve toked in the last few hours?

That’s the vision of Hound Labs, a Bay Area company that announced Wednesday it had completed a successful clinical trial for its “world’s first” marijuana breathalyzer. The company says it will move forward with making a commercial product with a roughly $5,000 price tag, primarily aimed at police and construction companies. 

Hound Labs claims the trial shows THC can be detected in the breath, in the truly minuscule amount of a trillionth of a gram per liter, for two to three hours after inhaling. This amounts to a “significant breakthrough that validates our science and technology,” says cofounder and CEO Dr. Mike Lynn, a former ER doctor and VC. 

Whether that is in fact what the trial says or does, we can’t say. The study has not yet been peer-reviewed or published. It asked 20 people — a small number for a clinical trial — to bring their own weed. The UCSF doctor responsible for the study says “financial restrictions” prevented her from using more participants, or a control group, but that they were not necessary.

“No study for the evaluation of THC in breath has been larger to date,” explains Dr. Kara Lynch. “Future studies will be done involving more participants and with varying experimental designs.”   

Mashable was shown a non-working prototype of the Hound, and several larger previous prototypes that supposedly proved the technology needed to detect that tiny amount of THC is getting smaller. But Hound Labs is not letting journalists — or indeed anyone, outside the 20 people UCSF pulled off the street — test the device at the moment. (In 2016, an early version was apparently tested on volunteer drivers in California. No high people were found.)

“We’re not Theranos,” joked Lynn, referring to the defunct Silicon Valley company that famously hyped a breakthrough blood test requiring minuscule amounts of blood. At the same time he admitted that even he had not read the UCSF study, and defended his policy of not demoing the product. “When it’s commercially available,” Lynn told us, “you can test it.”

Had we been able to try it, Lynn did admit, the test would show some THC in your breath no matter how much was in your system. A brief toke on a vape pen within the past few hours — or even a small bite of an edible — would show up on the Hound as clearly as a dozen bong rips. 

A breakthrough in detection, perhaps. But it raises the specter of someone getting fired or arrested for doing the cannabis equivalent of drinking half a beer. In a world that’s coming around to the medical benefits of marijuana microdosing, that’s a real problem.

We’d love to tell you exactly what and how those 20 participants ingested to get their results, but not even the UCSF doctor knows, thanks to restrictions on marijuana research. Hound Labs has no such restrictions, but also doesn’t seem to appreciate the world of difference between dose levels. 

Take this video of a test of stoned driving that Hound Labs conducted two years ago. Drivers — we’re not told how many participated — were told to ingest as much weed as they felt like. We see someone smoking what appears to be an entire joint. They then had to maintain speeds of up to 65 mph around a 1.5 mile track. Not surprisingly, they hit stuff. Lynn claims this is the equivalent of stoners hitting cyclists in the real world.  

Hound Labs is not the only company to be chasing the weed breathalyzer vision; a Canadian company called Cannabix is hot on the Hound’s tail. That might explain why Hound Labs was eager to claim its clinical trial was successful before verifiable results were in. It’s hardly the first company to grasp early for the laurel wreath of proof. 

At the same time, both companies are being hounded by some pretty essential questions: No matter the level of breakthrough, is a weed breathalyzer even necessary? Does it prove anything we didn’t know before? Does pot really need to be singled out like this? 

“It’s illegal to drive under the influence of many drugs,” says Paul Armentano, deputy director of the marijuana legalization advocacy group NORML. He points out that police have a number of sobriety tests already at their disposal to determine whether someone is too stoned or otherwise drugged out to drive, without having to resort to a breath or blood or urine sample. 

If you can’t walk in a straight line, if your eyes are bloodshot and your speech is messed up, that is already legally admissible evidence that you should have found a designated driver, dude.    

“Could [the Hound] be one piece of evidence? Of course,” admits Armentano. But he says Hound Labs is missing the bigger picture. For all the company’s 65 mph closed-course, whole-joint testing, “there is no science correlating the detection of THC in breath with impairment of performance.” 

He’s not wrong. A major 2015 U.S. Department of Transportation study found that “after adjusting for age, gender, race and alcohol use, drivers who tested positive for marijuana were no more likely to crash than those who had not used any drugs or alcohol prior to driving.” The same year a University of Iowa study showed drivers were safer behind the wheel after they smoked compared to after they drank. 

In 2013, an analysis of 66 studies found a slightly elevated risk of traffic-related injuries with marijuana use — but it was lower than the elevated risk from taking penicillin, antidepressants, and even antihistamines. 

You’re not exactly the best driver in the world when you’re high, but you will often err on the side of caution. A Harvard researcher once offered this memorable if exaggerated explanation: When you’re drunk, you run red lights. When you’re stoned, you stop at green lights. 

There is no agreement on an impairment level for THC anywhere else in the world, either. Canada can’t figure it out. A UK government transport report said that impairment is a thing, but lasts for just one hour after pot consumption — far short of the two to three hours Hound Labs is measuring. 

None of that has stopped a handful of states from imposing an arbitrary THC limit regardless. The fear of marijuana proponents is that the folks in charge will blindly trust detection technology like Hound Labs’, and impose new laws on that basis, without thinking through the implications. “Police, pundits and lawmakers already conflate breathalyzers with impairment,” says Armentano.

Even if Hound Labs’ tech is for real, even if it can  show exactly how much THC entered your system in the last couple of hours, it’s easy to see scenarios where cops might abuse it for an easy arrest of a non-impaired driver, or employers might use it to safely get rid of the long-haired guy they just don’t like. 

You’re not paranoid, stoners. The Hound may soon be out to get you.