Where NetSenX uses AI, and where it does not.
Article 50 of the EU AI Act (Regulation (EU) 2024/1689) sets transparency obligations for certain AI systems. NetSenX publishes this disclosure for every AI component in the product, including those the article may not require it for.
Section 1
Where AI is used in NetSenX
Three components use AI. Each has a documented scope, and where it applies, a setting you control.
Detection engine
Detection · always on
Scores how each host behaves on the network, on a continuous risk scale rather than a yes or no.
- The method behind the false-positive bound, split-conformal prediction at a stated confidence level, is published, so the construction can be checked independently. The bound in force for your tenant, and the calibration size behind it, are shown in the product.
- Each alert states its confidence in plain words and is labelled preliminary when its score is not calibrated.
- Each alert lists, in plain language, the network features that drove its score, on every plan.
NetSenX AI Analyst Core
Explanation · off until you set it up
Writes plain-language analyst notes for alerts the detection engine has already raised.
- It is off until you set it up under Settings › AI Analyst with your own AI provider key, on any paid plan. You can switch it off there at any time, and detection continues without it.
- Where a provider processes data outside the EU, the product asks for your explicit consent before using it.
- It does not take response actions. Automated responses run only from playbooks your team writes or copies from NetSenX templates.
Section 2
What NetSenX does not use AI for
Stated here so customers can rely on it in their own risk assessments.
Automated responses decided by AI
Automated responses run only from playbooks a customer writes or copies from NetSenX templates. A new playbook starts as a draft and does nothing until it is armed. Neither the AI Analyst Core nor any language model starts a response.
Section 3
What you can do
Each item below is a product feature, not a request form.
See why
Every alert comes with a plain-language explanation of the network features behind its score, on every plan.
Switch it off
Switch the AI Analyst Core off under Settings › AI Analyst at any time. Detection continues without it.
Take your data
From the Pro plan, which includes REST API access, you can retrieve your alerts as JSON through the API.
Audit what happened
Every AI-derived decision is recorded in an append-only trace that the database refuses to change or delete, available via dashboard and API. The administrative audit log is kept for the period your plan or your own retention policy sets, never less than 90 days, and is then moved to cold storage rather than deleted.
Section 4
Risk classification under the AI Act
Our classification
We assess NetSenX as a limited-risk AI system under the AI Act: it informs human decisions and does not take legally significant decisions on its own. The transparency obligations of Article 50 apply.
What we apply voluntarily
For the AI Analyst Core we apply, voluntarily, practices the AI Act sets for high-risk systems: decisions are logged and every note is shown with the model version that wrote it.
If you run NetSenX inside a critical-infrastructure or public-administration workflow, your own use may fall under stricter rules. Ask us for the documentation your assessment needs.
Section 5
Governance and ISO/IEC 42001
Status: in preparation, no certification yetWe are building an AI management system aligned with ISO/IEC 42001. It has no certification yet, and no audit date is set. Model versions are pinned and recorded, so each note and alert can be traced to the model that produced it.
Section 6 — Contact
AI-related questions go to the DPO.
The Data Protection Officer is also the owner of AI accountability: one contact for questions about any AI component on this page.
dpo@netsenx.com