AI Disclosure — K0nsult

K0nsult provides exclusively independent audits of AI systems. All audits are currently provided free of charge during the deployment pilot phase. This page explains, in plain language, which AI models we use, what those models do and do not do, how AI-generated content is labelled, and how a human stays in control of every output that affects you.

1. Which AI Models We Use

K0nsult uses large language models supplied by Anthropic. We do not train our own foundation models and we do not allow our providers to train on customer data.

Model	Identifier	Where it is used
Anthropic Claude Haiku 4.5	`claude-haiku-4-5-20251001`	The Lyra assistant — message triage, classification, and first-line user support
Anthropic Claude (Sonnet / Opus class)	Selected per task	Optional — preliminary audit analysis and drafting support, always reviewed by a human auditor

If we add or change a model, this page will be updated and the change will be reflected in the metadata attached to AI-generated content (see Section 4).

2. What AI Does

We use AI strictly as an assistive tool. Specifically, AI helps with:

Message classification / triage — sorting and prioritising incoming messages so they reach the right person quickly
User support — answering common questions and guiding users to the right resource
Preliminary audit analysis — drafting initial observations, structuring evidence, and surfacing potential findings for a human auditor to review

3. What AI Does NOT Do

AI at K0nsult never acts autonomously where it matters. AI does not:

Take autonomous decisions that affect a person's legal rights or have a similarly significant effect
Execute or authorise financial transactions
Provide binding legal, medical, or regulatory advice without human review and sign-off
Finalise or issue an audit deliverable without a human auditor's review and approval
Profile individuals to make solely automated decisions about them

4. How We Label AI-Generated Content

Any content produced or substantially assisted by AI carries a clear, visible label and machine-readable metadata so the recipient always knows when AI was involved.

Visible label: "🤖 AI-generated" shown next to the content
Metadata attached to the content:
- model — the exact model identifier used (e.g. claude-haiku-4-5-20251001)
- timestamp — date and time of generation (UTC)
- audit_log_id — reference to the immutable audit-log entry for traceability

5. Human Oversight (EU AI Act Art. 14)

A human is always in the loop for anything that affects a client. Our oversight controls are:

Human review: any AI output that affects a client — including every audit deliverable — is reviewed and approved by a human auditor before it is acted upon or delivered
Kill switch: AI assistance can be disabled immediately by a human operator at any time; the service continues to operate with humans only
Escalation contact: oversight and escalation requests go to kontakt@k0nsult.cloud

No decision affecting your rights is fully automated. The final responsibility for every audit deliverable and every decision rests with a human auditor. This is consistent with Art. 14 of the EU AI Act (human oversight of high-risk AI systems) and Art. 22 GDPR (automated individual decision-making).

6. Your Rights

When you interact with K0nsult, you have the following rights regarding our use of AI:

Right to be informed: you have the right to know when you are interacting with an AI system or receiving AI-generated content
Right to a non-AI alternative: you may request that your matter be handled by a human without AI assistance
Right to an explanation: you may request an explanation of the role AI played in any output that concerns you, including the model used and the human review applied

To exercise any of these rights, contact us at kontakt@k0nsult.cloud.

7. Legal Basis

This disclosure is provided to satisfy our transparency obligations under:

EU AI Act, Art. 13 — transparency and provision of information to deployers
EU AI Act, Art. 50 — transparency obligations towards natural persons interacting with AI systems and receiving AI-generated content
EU AI Act, Art. 14 — human oversight (referenced in Section 5)
GDPR, Art. 22 — the right not to be subject to a decision based solely on automated processing, including profiling

8. Contact — Data Protection Officer (DPO)

For any question about our use of AI, AI-generated content, or to exercise your rights:

K0nsult / CODE NO CODE — Data Protection Officer (DPO)
Email: kontakt@k0nsult.cloud

See also our Privacy Policy and Trust Center for related disclosures.