Back
    Service

    AI & Compliance

    We help organisations use AI effectively and responsibly — with clear governance structures, concrete guardrails and a compliance strategy that does not block innovation but secures it.

    Resilience through AI requires governance

    AI without governance increases dependency and risk. Resilience only emerges once it is clear:

    • Who decides – and where a person gives final approval.
    • Which data the AI uses, where it is stored and who can access it.
    • How errors are detected, escalated and corrected.
    • Which requirements apply under the EU AI Act, NIS2 and GDPR.

    Governance is not a brake but the condition under which AI works reliably when it matters.

    Traceable and tender-ready: Documented roles, approvals and controls provide the kind of evidence clients may require in tenders.

    Concrete consulting modules

    Compliance diagnostics

    We analyse existing AI initiatives, responsibilities, policies and risk areas — from data protection to model usage.

    Governance design

    We define roles, decision rights, approvals, documentation obligations and guardrails for secure AI use.

    Use case prioritisation

    We assess AI use cases by impact, risk, regulatory proximity and operational feasibility.

    Controls & escalation

    We develop monitoring, review cycles and response logic for breaches, misconduct and undesired AI effects.

    Governance model for AI

    Our model combines accountability, risk management and operational steering. AI becomes controllable, traceable and scalable — without losing its impact.

    Preventive

    Define guardrails before deployment: roles, policies, data standards, approvals and no-go zones.

    Proactive

    Recognise and prioritise regulatory changes, sensitive processes and critical risks at an early stage.

    Detective

    Make usage, outputs, access and deviations visible through reviews, audits and monitoring.

    Reactive

    Act quickly on incidents: secure escalation, correction, communication and learning organisationally.

    What the AI Act, NIS2 and GDPR actually require

    EU AI Act

    • AI literacy: organisations using AI must ensure their staff are sufficiently trained (Art. 4, applies since 2 February 2025; as of: November 2026).
    • Prohibited practices: certain AI applications, such as social scoring or emotion recognition in the workplace, are banned (Art. 5, applies since 2 February 2025; as of: November 2026).
    • Transparency: people must be able to recognise that they are interacting with an AI system; AI-generated content must be labelled under certain conditions (Art. 50).
    • High-risk systems, e.g. in recruitment: deployers must ensure human oversight, retain logs and use the system according to its instructions (Art. 26). Following the Digital Omnibus, these obligations apply to high-risk AI under Annex III (e.g. HR, credit scoring, critical infrastructure) from 2 December 2027 and under Annex I (embedded in regulated products) from 2 August 2028 (as of: November 2026).

    NIS2 (in Germany: NIS2 Implementation Act)

    • Check applicability: obligations apply to companies in certain sectors above a size threshold; affected entities must register with the BSI.
    • Risk management: appropriate technical and organisational measures, including access control, backup, incident management and supply-chain security — including the AI services you use.
    • Reporting: significant incidents as an early warning within 24 hours, a notification within 72 hours and a final report after one month.
    • Management: it must approve the measures, oversee their implementation and attend training regularly.

    GDPR

    • Legal basis: every processing of personal data by AI requires a legal basis (Art. 6).
    • Processors: a data processing agreement is required with external AI providers (Art. 28); transfers to third countries need additional safeguards.
    • Data protection impact assessment: required before deployment where a high risk to data subjects is likely (Art. 35).
    • Automated decisions: decisions with legal or similarly significant effects may in principle not be based solely on automated processing (Art. 22); data subjects must be informed (Art. 13/14).

    Note: This overview is for orientation only and does not constitute legal advice. Whether and to what extent the rules apply to your organisation should be clarified with your legal counsel or data protection officer.

    As of: November 2026

    Typical use cases

    Generative AI in knowledge access

    Guardrails for internal assistants, sensitive content, source control and human approvals.

    Document & contract review

    Governance for accuracy, traceability, data protection and accountability hand-offs in approval processes.

    AI in sensitive processes

    Control mechanisms for HR, finance or compliance-adjacent decisions with high risk exposure.

    Leadership & enablement

    Training for executives and teams so AI is used not only technically but also in compliance with the rules.

    Secure AI effectively

    If you want not only to introduce AI but to govern, prioritise and scale it in a structured, compliant way, we will develop a robust target picture and the matching implementation modules with you.