Workshops: EU AI Act & Compliance

Information Security & AI under ISO 27001: Workshop

AI tools change the threat landscape and need to be part of your ISMS. This workshop shows security and ISMS owners how to cover generative AI and AI agents in risk assessment, controls and supplier management under ISO 27001.

Who is this workshop for?

  • ISMS owners and information security officers
  • CISOs and security teams
  • IT architecture and operations
  • Internal auditors
  • Organisations preparing for ISO 27001 audits

What you will take away

  • Add AI-specific threats such as prompt injection, data leakage and over-privileged agents to your risk assessment.
  • Know which ISO 27001 Annex A controls matter most for AI use.
  • Assess AI vendors systematically in supplier management.
  • Apply technical safeguards for AI assistants, copilots and coding agents.
  • Know what evidence auditors typically expect on AI topics.

Agenda

  1. Module 1

    AI and the threat landscape

    • New attack surfaces: prompt injection, data poisoning, abuse of tool access
    • Data leakage through chat inputs and uploads
    • AI on the attacker side: phishing and social engineering
  2. Module 2

    Anchoring AI in the ISMS

    • Extending the asset inventory with AI services and models
    • Risk assessment for AI applications
    • Relevant controls: supplier relationships, cloud services, access control, data masking, logging
  3. Module 3

    Suppliers and cloud AI

    • Assessing ChatGPT Enterprise, Claude for Work, Microsoft 365 Copilot, Google Gemini and Mistral
    • Data residency, certifications, retention and training use
    • Minimum contractual requirements
  4. Module 4

    Technical safeguards

    • Permission models for copilots and agents with MCP access
    • DLP, sensitivity labels and network controls
    • Sandboxing and approvals for coding agents like Claude Code or GitHub Copilot
  5. Module 5

    Audit preparation

    • Typical auditor questions on AI
    • Evidence and documentation
    • Bridging to ISO/IEC 42001 and the EU AI Act

Why AI is an ISMS topic

Every AI tool that processes company data is both an information asset and a supplier. Copilots with access to email and documents also amplify existing permission mistakes: what used to be hard to find, the AI surfaces in seconds.

Agentic systems that take actions add another dimension: manipulated content can trick an agent into sharing data or running commands. These risks belong in the risk assessment.

First-hand experience

jocapps, the company behind neooutput, runs an ISO 27001 and ISO 9001 certified management system and uses AI extensively itself, including in software development. Our AI Trainers and AI Code Engineers share what works in practice.

  • Field-tested risk scenarios
  • Examples of controls and evidence
  • Realistic assessment rather than alarmism

Format

The workshop runs remotely or on-site, in English or German. We work with your existing ISMS and the AI tools you actually use.

Frequently asked questions

Does ISO 27001 need AI-specific controls?

The standard has no AI-specific controls, but many existing ones apply directly, such as those for suppliers, cloud services and access control. We show how to apply them to AI and where additional measures make sense.

What is the biggest risk with Microsoft 365 Copilot?

Often it is overly broad permissions in SharePoint and Teams that Copilot makes visible. Cleaning up permissions before rollout is therefore essential.

Do you cover coding agents?

Yes. Agents that execute code and access repositories are a risk area of their own. We discuss sandboxing, permissions and approval workflows.

Workshop

Get in touch

Shall we integrate AI cleanly into your ISMS together?

Tell us a little about your team – we will get back to you within 24 hours with a proposal for content, format and dates.

Prefer to talk?

We only use your details to handle this request. See our privacy policy.