Report an Incident

Notice-and-action channel under DSA Art. 16, EU AI Act Art. 50/85, NetzDG · v2026-04-18

Emergency / acute danger: If someone is in immediate danger, call your local emergency number (Germany: 110 police / 112 medical) or equivalent. This page is NOT a substitute for emergency services.

Zeptix operates a platform on which third parties (so-called „bot operators" or „tenants") publish their own AI chatbots. If something goes wrong while you use one of those bots, you have several reporting channels.

1. Which channel should I use?

1.1 Inside the chat: Every bot reply has a „report" button. The report goes straight to the bot operator, who must process it within 48 hours (24 hours for harmful / illegal, immediately for critical cases — see Terms § 7a).

1.2 Direct to the bot operator: Their contact details are in the bot’s imprint/profile, linked from the chat page. If the operator has not provided their own imprint, contact us instead (1.3).

1.3 To Zeptix as platform provider: If the operator does not respond or it is a systemic problem (hosting abuse, illegal content, CSAM, terror), email us at [email protected]

2. What should I report?

CSAM — zero tolerance, removed immediately, reported to BKA/NCMEC.
Self-harm — suicide instructions, dangerous diet advice, pro-ana, self-injury.
Violence & terror — weapons/bomb instructions, attack planning, incitement.
Illegal content — drug synthesis, hacking guides, counterfeit goods, copyright violations.
Hate speech — slurs based on origin, religion, gender, sexual orientation, disability.
Privacy violations — bot exposes or stores personal data without legal basis.
Defamation — false factual claims about identifiable persons.
Spam, scam, fraud — phishing, investment scams, crypto pumps.

3. What happens with my report?

We confirm receipt typically within 24 hours. For reports to [email protected] we review the case against our Acceptable Use Policy:

No violation — we reply with a brief explanation, the bot stays online.
End-user violation — forwarded to the bot operator, who can suspend or ban the user.
Operator violation — we warn, throttle, suspend or remove the bot (graduated sanctions, see AUP § 6).
Criminally relevant — we preserve evidence and notify law enforcement or specialized bodies (NCMEC, INHOPE, Europol IRU).

4. Required information for a valid notice (DSA Art. 16)

To help us act quickly, please include:

URL of the affected bot (e.g. chat.zeptix.io/en/<tenant>)
Date/time of the incident (as precise as possible)
Exact description of the problematic content (quote or screenshot)
Why it is a violation (category from 2., or applicable law)
Your contact details for follow-up (anonymous is possible but slower)
A statement that your information is accurate to the best of your knowledge (DSA requirement)

5. Privacy-specific concerns (GDPR)

For data protection issues (e.g. a bot displays your personal data without consent), email [email protected]. Note: in many cases the bot operator is the „controller" within the meaning of Art. 4(7) GDPR (see joint-controllership notes). We will forward your request to the responsible party, or you can contact the address listed in the bot’s imprint directly.

You may also lodge a complaint with the competent supervisory authority at any time (Art. 77 GDPR).

6. EU AI Act — complaints about AI systems (Art. 85)

From 2 August 2026, natural persons may file a complaint under Art. 85 EU AI Act with the competent market surveillance authority regarding allegedly prohibited or high-risk AI practices. Zeptix will provide logs, training-data information and configurations to such authorities to the extent required by law.

7. Authorities, law enforcement, researchers

Requests from law enforcement, supervisory authorities or accredited researchers (DSA Art. 40) should go to [email protected]. We also accept service of process by mail to the address in our imprint.

8. Protection against abusive notices

Knowingly false notices may have civil and criminal consequences (DSA Art. 23(2) and applicable national law). We log all reports and may suspend repeat false reporters.

This page is updated regularly. We bump the version number at the top for legally relevant changes. Last update: 2026-04-18.