Free handbook · 42 pages · Marram 2026

AI is already working in your area. Who is responsible for its quality?

AI assistants and agents already do HR, quality and communications work. They use your organisation’s guidelines, interpret information and prepare decisions. The Context Owner’s Handbook shows how to make sure they apply your company’s approved principles correctly, and where to start.

We will send the download links to your email. We may also send you Marram content related to your area, and you can unsubscribe from the link in every message. Privacy notice

Practical examples · Worksheets · PDF, Markdown and the Context Check skill

Cover and inside pages of The Context Owner’s Handbook
Written for leaders responsible forPeople & Culture · Quality · Communications · Sales · Customer service · Projects · Sustainability
ExamplesKONE · Ericsson · Bosch · IKEA · KlarnaResearchDeloitte · Boston Consulting Group · Bain & Company

A new kind of handbook

One handbook, three formats.

PDF for peopleEasy to read and to share with your leadership team.
Markdown for AIGive the handbook to your own AI assistant and ask it questions from the perspective of your area.
Context Check skillTest your own guideline: the skill shows how AI interprets it and what is missing.ClaudeChatGPT

Why now

Your area has new workers.

The quality director is still responsible for quality, the HR director for HR principles and the communications director for the company’s communications.

AI tools now use the knowledge of their areas too, often tools they did not build or introduce themselves. This raises three questions:

  1. Where does AI get the organisation’s approved operating principles?
  2. Who makes sure it applies them correctly?
  3. How do we know the quality of its work is good enough?

These are not IT questions.

56%of Nordic organisations have given AI tools to at least 40 per cent of their staff.
16%have redesigned work extensively around them.
Deloitte, State of AI in the Nordics 2026 (170 leaders)

What the handbook gives you

Four things you can take to your leadership team.

1

A new perspective on your leadership responsibility

How the work AI does relates to the existing responsibilities, quality criteria and decisions of your area.

2

Where trust in AI’s work comes from

Why checking afterwards is not enough on its own, and how to build the conditions for reliable work in advance.

3

A model for shared context

Which knowledge is owned in common and what belongs in the instructions of an individual AI tool.

4

A practical way to start

How to choose your first body of content, identify the missing application guidance and define the tests.

Contents

Six chapters and four worksheets.

Download the free handbook
  1. 1People are no longer the only ones working in your area
  2. 2Where does trust in AI’s work come from?
  3. 3What does AI need beyond the guidelines?
  4. 4Shared principles, different AI tasks
  5. 5How do you prepare expert knowledge for AI use?
  6. 6Who owns this, and where do you start?
  7. A–DContext map · Five test situations · Self-assessment · What to record

Examples

Well-founded questions. Concrete examples.

The handbook looks at the public company guidelines of KONE, Ericsson, Bosch, IKEA and Klarna. The examples show why finding the right document does not yet mean that AI can apply it as intended.

It also draws on the perspectives of Deloitte, Bain and Boston Consulting Group on AI adoption, the role of expert knowledge and managing an organisation’s own operating context.

EricssonWhen the right information ends up in the wrong context
BoschA precise number does not always make an unambiguous rule
KONEHuman judgement cannot be transferred as it is

The examples are based on public material. They are not audits of these companies’ AI systems.

Who it is for

For the leader of an area, not for IT.

For the people responsible for HR, quality, communications, marketing, safety, procurement and other functions. Especially if the guidelines, decision rules and expertise of your area are used in AI solutions now or soon will be.

You do not need technical AI skills. The handbook is about the quality of work, shared operating principles and leadership.

The Context Owner’s Handbook

How can an organisation extend its expertise, decision rules and responsibilities to the work AI does?

A handbook with practical examples, worksheets and questions for assessing your own area.

You get

  • the handbook as a PDF for people and in Markdown for AI
  • the Context Check skill for Claude and ChatGPT

We will send the download links to your email. We may also send you Marram content related to your area, and you can unsubscribe from the link in every message. Privacy notice

Does the handbook cost anything?No. The handbook and the Context Check skill are free.
Do I need technical skills?No. The handbook is written for the leader of an area, not for IT. Installing the skill takes a few minutes.
What does Context Check do?It assesses how well one guideline you give it suits a chosen AI use and names the most important gaps. It does not test the behaviour of a real agent. The skill runs in your own Claude or ChatGPT environment, and your material is not sent to Marram.
Are the examples audits?No. They are based on the companies’ public guidelines and illustrate interpretation risks.
Why Markdown as well?You can give the handbook to your own AI assistant and ask it questions from the perspective of your area.
Can Marram do this with us?Yes. Library takes the core content of your area to levels 3 and 4 of the handbook and keeps it up to date in all your agents.

Publisher: Marram

Models change. Your judgement stays.