Guardrails First: Engineering Member-Facing Health AI — Rashi Agrawal, Hinge Health
A healthy 60 year old man asked a popular AI assistant how to cut salt from his diet. It pointed him at sodium bromide. Three months later he arrived in an emergency room with paranoia and hallucinations, bromide at 200 times the safe level, and stayed three weeks. Rashi Agrawal stacks that against the first independent safety test of a consumer health AI, out of Mount Sinai, which under triaged life threatening emergencies half the time, and against ECRI naming chatbot misuse the top health technology hazard of 2026. Roughly 40 million people already triage themselves this way. None of it is a frontier problem. It is the production baseline.Her argument is that most healthcare AI safety failures are architectural decisions made before a single token is generated. PHI is stripped at the pipeline boundary on ingestion, so a developer who opens a dashboard finds nothing to redact because it was never stored. Anything that can never be wrong lives in a code layer above the model rather than in its prompt: routing to 911 or 988, deciding which capability owns a turn, verifying who is on the other end. The frontier labs publish an authority hierarchy in which every layer above the user sits one prompt injection from being overridden, and her reading is blunt: if they will not treat a prompt as a security boundary, neither should you. Safety then runs as a continuous layer of judges scoring live traffic, with one discipline attached. When a score drops, first ask whether the judge is right.
Speaker info:
- https://www.linkedin.com/in/rashi283/
- https://sessionize.com/rashiagrawal/
Timestamps:
0:00 - The state of healthcare AI, and 40 million self triagers
1:04 - Poisoned by a chatbot
1:30 - Under triaging emergencies half the time
2:35 - Three non negotiable foundations
3:41 - Where PHI actually lives
5:53 - Deterministic rules belong above the model
7:27 - If the labs will not trust the prompt, neither should you
7:54 - Escalation, intent routing, identity
9:39 - Safety as a continuous evaluation layer
12:47 - Five stakeholders, five risks, five days to launch
14:02 - The five rules for deciding
18:10 - Verify the scorer before you trust the score
20:24 - The whole talk in one slide Receive SMS online on sms24.me
TubeReader video aggregator is a website that collects and organizes online videos from the YouTube source. Video aggregation is done for different purposes, and TubeReader take different approaches to achieve their purpose.
Our try to collect videos of high quality or interest for visitors to view; the collection may be made by editors or may be based on community votes.
Another method is to base the collection on those videos most viewed, either at the aggregator site or at various popular video hosting sites.
TubeReader site exists to allow users to collect their own sets of videos, for personal use as well as for browsing and viewing by others; TubeReader can develop online communities around video sharing.
Our site allow users to create a personalized video playlist, for personal use as well as for browsing and viewing by others.
@YouTubeReaderBot allows you to subscribe to Youtube channels.
By using @YouTubeReaderBot Bot you agree with YouTube Terms of Service.
Use the @YouTubeReaderBot telegram bot to be the first to be notified when new videos are released on your favorite channels.
Look for new videos or channels and share them with your friends.
You can start using our bot from this video, subscribe now to Guardrails First: Engineering Member-Facing Health AI — Rashi Agrawal, Hinge Health
What is YouTube?
YouTube is a free video sharing website that makes it easy to watch online videos. You can even create and upload your own videos to share with others. Originally created in 2005, YouTube is now one of the most popular sites on the Web, with visitors watching around 6 billion hours of video every month.
