New here? See Genox Cybersecurity AI in three minutes → Report an incident (24/7)
Trust Center

How we handle your data,
written down in plain terms

Which certifications we plan to get and when, where customer data is processed, and the rules our AI verdicts follow.
When any of this changes, we update this page first.

Certification roadmap

Certifications: none held yet

As of 2026.11.07 Everything below is in preparation or planned. We will not say "certified" until we are.

Security control design

Isolation architecture, access control and audit logging

ISO/IEC 27001 audit

Information security management system and internal audit

SOC 2 Type II observation period

Evidence of security and availability controls in operation

ISMS-P and ISO/IEC 42001

Korean management system certification and AI management system

Until certifications arrive, we answer security questionnaires and due diligence requests directly. Email admin@genox.one.

Data principles

How customer data is handled

Not used for training shared models

Customer data is never used to train shared models or to analyse other customers.

Where it is processed

On-premises installs process data inside your data centre. Dedicated cloud uses a separate tenant per customer in a Korean region.

Every transfer is logged

Each data transfer leaves a record that you can read yourself.

You set retention

You decide how long analysis data is kept and when it is deleted.

Staff access is limited

Genox staff need approval for each access to customer data. Access expires after a set time and is logged.

Korean data stays in Korea

Data of Korean customers is kept in Korea. If it ever has to move abroad, we ask for consent first.

Responsible AI

Rules our AI verdicts follow

No evidence, no verdict

Every verdict is recorded together with its evidence, reasoning and confidence.

Past verdicts can be re-run

We store the data, model version and policy version used for each verdict, so it can be re-run under the same conditions and compared. Wording may differ between runs. If the verdict itself changes, the difference is recorded.

People approve high-impact actions

Automatic response only runs inside the policy agreed with you. Isolation, blocking and account locks outside that policy need an analyst's approval.

False positives and misses are reported

Monthly customer reports for the full service will include false positive rates and missed cases. Before launch, we have published how we will measure them on the AI quality page.

Security architecture

How Genox itself is secured

Dedicated servers per customerAnalysis runs only on servers set aside for that customer
Network policyDeny by default, open only allowed connections, log every connection
EncryptionEncryption at rest and in transit, customer-managed keys (planned)
Audit logsSigned logs kept in tamper-proof storage
Responsible disclosure

Found a vulnerability in Genox?

Please tell us.
We will not take legal action against researchers who report in good faith and follow the rules here, and we will get back to you once we have checked it.

How to report

  • Email admin@genox.one with steps to reproduce and the impact.
  • Please allow at least 90 days for a fix before disclosure.
  • Do not run tests that access customer data or disrupt the service.
security.txt: /.well-known/security.txt

Talk to a Genox
security expert.

Talk to us