Brave Died. The Tech It Proved Didn't.

In January 2025, Canada shut down Brave, an app that let someone using drugs alone stay connected to a remote responder who could send help if they went silent. It didn’t fail. It ran out of money and institutional cover, and it died anyway. That’s the fact anyone excited about the next wave of overdose-prevention tech has to sit with before getting excited about the next wave of overdose-prevention tech.

Because there is a next wave, and it’s real. The technology-innovation beat has been thin lately, but not because nothing’s happening — it’s happening quietly, in commentary and peer-reviewed case studies rather than product launches. A recent Reason Foundation analysis argues digital tools are moving past novelty into actual infrastructure: predictive relapse-risk scoring built from engagement and check-in data, drug-checking tech, AI systems trained to flag at-risk patterns before an overdose happens.

Plain language on what that actually means: predictive relapse-risk scoring takes the digital exhaust of someone’s recovery — how often they log in, how they answer check-ins, whether their patterns change — and turns it into a risk number a clinician or peer supporter can act on. Virtual supervised consumption is the Here4UScotland model, studied in a peer-reviewed case report: someone uses drugs while connected by video or phone to a trained supervisor who can call emergency services the moment something looks wrong. No in-person site required. No one using completely alone.

THE TAKE: a relapse-risk score with no funded human staffed to respond to it isn’t harm reduction — it’s surveillance with a timestamp.

That’s not a hypothetical concern. It’s what Brave’s shutdown demonstrates directly. The app’s core function — a person on the other end of the line, ready to act — is also its most expensive, least fundable line item. Software is cheap to scale. Staffing a 24/7 response team that treats every alert as a life-or-death page is not, and it’s the part legislators and funders keep treating as optional.

Here4UScotland’s own case study is instructive precisely because it documents infrastructure, not a pilot with a press release. That’s the right bar. A relapse-risk model or a virtual supervision line is only as good as the funded, staffed, legally protected response behind the alert — the thing Brave had, and lost anyway. Anyone pitching this category owes a straight answer to one question: when the algorithm fires at 3 a.m. two years from now, who’s still funded to pick up?