Live Nov 18 · Cohort 3Nov 18, 19:00 MSK — free hands-on AI agents live workshop · Live Cohort 3 starts Nov 23Join Nov 18 Live →Cohort 3 (Nov 23)
ReleaseInsight · On RadarThe Verge→ original

Kintsugi shuts down: voice-based depression detection startup failed to win FDA approval

California startup Kintsugi is shutting down after seven years of developing AI that detected depression and anxiety from a person's voice. The company ran out of time to secure FDA approval and is now open-sourcing its code. Some of the technology may find use in detecting deepfake audio.

AI-processed from The Verge; edited by Hamidun News
Kintsugi shuts down: voice-based depression detection startup failed to win FDA approval
Source: The Verge. Collage: Hamidun News.
Listen to article

California startup Kintsugi is ceasing operations after seven years of developing artificial intelligence that can identify signs of depression and anxiety disorder from a patient's speech. The reason for the shutdown: it could not obtain approval from the US Food and Drug Administration (FDA) in time. Most of the technology will be published openly.

The problem Kintsugi set out to solve is real and large. Diagnosing mental health disorders still relies mainly on questionnaires and clinical interviews, unlike physical illnesses, where lab tests and imaging have long been standard. The startup took a different approach: rather than analyzing what a person says, it studied how they say it, including tempo, pauses, voice modulation and intonation patterns. According to the company's hypothesis, these acoustic markers correlate with depression and anxiety.

The technology looked promising: voice analysis can be done remotely, quickly and without special equipment. That opened up possibilities for initial screening in telemedicine, for monitoring patients between doctor visits, and for helping regions with an acute shortage of psychiatrists. The company raised investment and refined its algorithms for years.

However, medical regulation in the US is one of the hardest barriers for AI startups to clear. The FDA requires clinical evidence of efficacy and safety, and the approval process can drag on for years. For Kintsugi, the combination of regulatory timelines and, apparently, financial constraints proved critical.

Still, the story does not end in outright failure. Publishing the technology openly gives it a chance at a second life. According to The Verge, some of the work may be applied outside medicine, in particular to detect synthetic audio and deepfakes. This area is quickly gaining relevance as the quality of AI voice generation improves.

The Kintsugi case clearly illustrates a core contradiction in medical AI: the technology may work, but few can prove it to a regulator within the required timeframe and on the required budget. Voice-based diagnosis of mental health disorders as a field will not disappear. Other players will keep developing it, including by building on Kintsugi's now-open work.

Story
ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Unified AI Workspace · Nomad

Test Frontier LLMs & Autonomous Agents in One Workspace

Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.

Try Nomad AI →Launch Nomad Agent
What do you think?
Loading comments…