Raw DNA upload privacy

Is it safe to upload DNA data? What to check first

Uploading raw DNA data is never risk-free. Before using any service, check whether the full file leaves your device, what derived genetic data remains, who can access it, whether it is used for research or advertising, and how deletion works. DNA Layer reads the original locally and does not intentionally retain that full file.

Published by DNA Layer · Product behavior and public sources reviewed 26 August 2026 · DNA Layer is described as a first party · No independent security audit or medical review is claimed

The question is not just “is the upload encrypted?”

Encryption and access controls matter, but the larger privacy decision spans the whole data lifecycle. NHGRI notes that once you download genetic information and give it to another service, the original testing company is no longer responsible for that copy. NIST likewise identifies privacy and cybersecurity gaps across the generation, processing, sharing, storage, and aggregation lifecycle for genomic data.

What happens in DNA Layer’s current upload flow

Stage	What happens	What remains sensitive
Original file	The browser reads the raw export on your device. DNA Layer does not intentionally retain that complete file in this flow.	Your local copy still needs secure storage. Do not email it casually or upload another person’s file without authority.
Supported markers	Supported marker data and technical metadata needed to build the layer are sent for processing.	This filtered material is still genetic data; local reading does not mean that nothing genetic crosses the service boundary.
Saved layer	Supported markers, evidence records, generated findings, and guidance are stored privately so you can return to the layer.	Derived findings can reveal health-related information even without the original file.
AI-assisted explanation	Automated systems and specialist infrastructure help organize evidence and draft explanations, with deterministic evidence and safety checks around the synthesis.	Generated guidance can be incomplete or wrong and is not anonymous merely because it is derived.
Analytics	PostHog and Amplitude receive limited journey and technical events. Automatic interaction capture and session replay are disabled.	DNA Layer’s policy excludes email, DNA, genotype, health-report content, authentication secrets, and card or bank details from those analytics events.

These are the current public commitments, not an independent audit. Read the full DNA Layer privacy policy before uploading.

DNA Layer’s raw-file, derived-data, access, and deletion commitments

Original-file flow	Report or derived data	Material access	Deletion qualification
Read locally in the browser; the original full file is not intentionally retained by DNA Layer. Original raw file not retained in the browser-based upload flow.	Supported marker data, evidence records, and generated findings are stored privately for account access.	Specialist service providers may process data to operate the service; identifiable genetic data is not shared with advertisers or data brokers.	Deletion is available by request; data no longer needed is deleted or irreversibly de-identified, subject to legal needs and normal backup expiry.

The same source row powers the public DNA-analysis privacy matrix, which compares published documentation for seven services without declaring an overall winner.

Seven questions to answer before any DNA upload

Does the complete file leave my device? “Processed securely” does not answer where processing happens.
What is retained? Separate the raw file, supported markers, generated report, account details, and backups.
Who gets material access? Check staff, infrastructure providers, laboratories, marketplace apps, and AI services.
What secondary uses are allowed? Look separately for research, advertising, sale, model training, and law-enforcement language.
Can I delete both the account and genetic artifacts? Note legal holds, completed research, and backup-expiry exceptions.
What happens after a merger, sale, or closure? A privacy policy should describe business-transfer conditions.
Is the output medically bounded? Strong privacy cannot make an incomplete consumer array diagnostic.

Use the full private DNA-analysis checklistwhen comparing services. The FTC similarly advises checking who can access genetic information and how a company protects and uses it. 23andMe’s own guidance tells customers to review a third party’s privacy, security, sharing, and scientific-validity claims before uploading.

Why “de-identified” does not automatically mean anonymous

Genetic data is unusually persistent: unlike a password, it cannot be replaced. The UK Information Commissioner’s Office explains that a genetic record can remain identifiable when enough distinctive markers are present, even after names and obvious identifiers are removed. Treat partial marker data and derived health findings as sensitive unless a service explains a defensible anonymization boundary.

Privacy and medical accuracy are separate decisions

A private service can still overinterpret a file, and an evidence-led service can still have a data flow that does not suit your risk tolerance. Consumer genotyping covers a selected marker set rather than every variant. Important findings may require confirmation in an appropriate clinical laboratory before treatment, medication, screening, or reproductive decisions.

See how DNA Layer matches evidence and applies safety gates, or read why a general chatbot is not a raw-DNA pipeline.

A safer upload routine

1
Keep one untouched original
Store the provider ZIP privately and avoid opening or resaving the marker table in spreadsheet software.
2
Verify the service and file
Use the service’s official domain, confirm the supported provider and format, and do not upload a PDF, screenshot, family tree, or someone else’s DNA.
3
Read the current policy
Check full-file handling, derived-data retention, access, secondary use, deletion, and business-transfer terms before choosing the file.
4
Review what was saved
After processing, inspect the account controls and retain only the outputs you actually need.

Frequently asked questions

Is it safe to upload raw DNA data?

No DNA upload is completely risk-free. Check whether the whole file leaves your device, what is retained after processing, who can access or reuse it, whether data reaches third-party apps or AI providers, and how deletion works. A local-read design reduces exposure but does not make derived findings non-sensitive.

Does DNA Layer store my original raw DNA file?

In DNA Layer’s current browser-based upload flow, the original file is read on your device and is not intentionally retained. Supported marker data and technical metadata needed to build the layer cross the service boundary, and derived findings are stored privately for account access.

What genetic information does DNA Layer keep?

DNA Layer stores supported marker data, evidence records, generated findings, and guidance needed to provide your saved layer. Those records are derived from the file and remain sensitive personal data even though they are not the original full export.

Is my raw DNA file sent to an AI model?

DNA Layer’s browser flow filters the original file before server processing. AI-assisted systems and specialist providers help organize evidence and draft explanations from the supported data needed for that work; the original full file is not intentionally retained. Generated guidance is safety-gated but can still be incomplete or wrong.

Does DNA Layer send genetic data to analytics companies?

No. DNA Layer’s privacy policy says PostHog and Amplitude receive limited journey events, not your email, DNA, genotype, health-report content, authentication secrets, or card or bank details. Automatic interaction capture and session replay are disabled.

Can I delete my DNA Layer data?

You may request account deletion. DNA Layer deletes or irreversibly de-identifies data it no longer needs, except where retention is required by law or needed for legal claims; secure backups expire on their normal cycle. This is not a promise of instantaneous deletion from every backup.

Is de-identified DNA data anonymous?

Not automatically. Genetic records can remain identifying when enough distinctive markers are present. The UK Information Commissioner’s Office treats identifiable genetic data as special-category data even when names or obvious identifiers have been removed.

Sources and public policies

Ready to inspect your file with the original kept off our servers?

Choose your DNA provider