Trust & safety
So here is exactly what Antelic does with it, exactly what it will never do with it, and exactly what stands behind the answers it gives you. Written to be checked, not to be taken on faith.
Four commitments, each built into how the product works rather than into a policy document. Breaking any one of them would be visible from the outside.
No shop, no marketplace, no affiliate links, no revenue share with any brand it mentions — not hidden, not planned. The moment a company earns money from what it recommends, every warning it gives you is worth less. When Antelic charges, it charges for the software. Never for what you take.
The safety layer reads your saved results and your stack directly, with no AI in the loop and no feature flag in front of it. Every account sees the same warning, at every price, including free. When it fires, it says once and plainly what it saw and tells you to contact a clinician. It cannot be switched off, upsold, or overruled by AI text.
No advertising tools, no analytics tools, no data brokers, no third-party trackers — none of them were ever installed, so there is nothing here to configure or to switch off. One crash reporter, handed an account id and nothing else: never a compound, never a marker, never a value. Nothing you save is used to train any model, here or anywhere else.
Every answer is grounded in a curated clinical knowledge base rather than in what a model remembered, and when Antelic interprets one of your lab results it cites the sources behind the range it judged the value against. Where the evidence genuinely is thin, Antelic says so in plain words. A confident answer that hasn’t been earned is worse than no answer, because you would have no way to tell the difference.
Antelic organizes and explains. Clinicians diagnose and prescribe. Antelic is not a medical device, it does not diagnose, it does not prescribe, and it never tells you what dose to take. When something looks urgent it says so and points you at a clinician — that is a reason to call your clinician, not a diagnosis.
What stands behind an answer
Antelic maintains a curated clinical knowledge base covering the markers it interprets, the compounds it identifies, the dosing units it displays, and the interactions it flags. Reference ranges, unit conventions, interaction and pairing rules and compound records ship with the product as reviewed reference files — not as something a model remembered. The ambition is complete coverage of everything Antelic speaks about, and it grows continuously.
When Antelic answers a question, it answers from that base. Where it is reading one of your own lab results, the interpretation carries the sources behind the ranges it was judged against, so you can tell reviewed reference material from general knowledge instead of having to guess which one you are looking at.
And where the published evidence on something genuinely is thin, Antelic tells you that instead of producing a fluent answer to fill the space. That is the hardest thing to build into a medical AI and the most valuable thing it can do: an honest the evidence here is limited is a usable answer. A confident guess is not.
Out of scope, and named here rather than buried in consent text: diagnosis, triage, ranking one brand against another, and any question whose honest answer is a quantity.
Data handling
In plain language, in order. This runs alongside the privacy policy rather than replacing it — the policy is with counsel and will be linked in the footer as soon as it is final.
Read
The image or PDF goes to the model that extracts values from it, and nowhere else.
Shown to you
The values come back into a form for you to see. Nothing has been saved at this point, and closing the form discards it.
Saved
Only what you confirm. It is written under your account id, and the rule that makes it unreadable by anyone else lives in the database itself, not in application code.
Discarded
The file. It is never stored, so there is no archive of other people’s lab reports to be breached.
Never cached
The extraction. Reference answers about public compounds are cached globally because they are identical for everyone; the contents of your report are not, and there is nowhere for them to be kept, because they are never collected for reuse in the first place.
Deleted
Whenever you want, from inside the product. Account and records together, without writing to anyone.
Provenance
Almost nobody in consumer health publishes this. It is the clearest way to show what an answer is actually made of.
What you take The name of each item, normalized against the reviewed unit and compound database that ships with the product.
Your uploaded file Once the values are extracted, it is gone. Not stored, not passed on.
Your saved values The marker values you confirmed, with the reference ranges that arrived with them.
Anything belonging to anyone else There is no cross-account data in any request. Nobody’s record is context for anybody else’s answer.
Your own logs The doses and symptoms you recorded, when you ask for a read of them.
Permission to name a dose Antelic is never asked what you should take, and the answers it is allowed to give have no room for a quantity.
Curated rules The interaction and pairing rules that ship with the product, as reviewed files rather than model memory.
Urgency Anything that looks urgent is handled by the deterministic safety layer with no model involved at all. The AI can neither raise a flag nor lower one.
The ceiling
The strongest word Antelic is permitted to use about a correlation is suggestive, and that ceiling is enforced in the code rather than written into a footnote. Where curated review exists, the answer shows it. Where it doesn’t, the honest answer is that the evidence is limited — and that is the answer you get.
Security
Where Antelic stands
Antelic is a health and wellness product, not a medical device, and it is deliberately built to stay on that side of the line: it organizes, explains and prepares you for a conversation with a clinician. Diagnosis, prescription and dose are a clinician’s work, and Antelic does not do them.
You will not find a compliance badge on this site, and here is why. There is no such thing as a HIPAA certification — no government body issues one. The seals that look like one are sold by private companies, and mean only that a private company sold a seal. Antelic applies the controls described above regardless of which framework ultimately governs it, and when counsel has settled the question, the answer appears in this paragraph in plain words.
Where we are
Everything described on this page is running today: records, stack and dose tracking, insights, the safety layer, the weekly read, the clinician document. Antelic is in early access, free while it is, and access is opening in stages so the product grows with the people using it rather than ahead of them. It runs in your browser, on a phone or a desktop.
Two things are in motion and named here rather than left to be discovered: independent clinical review of the safety layer is a committed step before general release, and the privacy policy and terms are with counsel and will be linked in the footer the moment they are final.
Disclosure
Write to security@antelic.com. Send what you found, how to reproduce it and what you think it exposes — no template, no form. You get a reply from a person, and Antelic will not pursue anyone acting in good faith who reports a problem to that address instead of using it. If a finding touches records that are not yours, please stop at the point of proof rather than going further.
