AI ADOPTION PLAYBOOK
Guide library
Tracks, worked examples, templates and methods: the complete guide, ready to read and use at your own pace.
Organisation tracks
- Organization-specific tracks
Adapt responsibilities and your first pilot to your situation.
- Independent professional track — 14 days
Adapt responsibilities and your first pilot to your situation.
- Micro-business / TPE track — 30 days
Adapt responsibilities and your first pilot to your situation.
- SME / PME track — 90 days
Adapt responsibilities and your first pilot to your situation.
- Nonprofit or foundation track — 60 days
Adapt responsibilities and your first pilot to your situation.
- Public-service track — formal gates
Adapt responsibilities and your first pilot to your situation.
Sector guidance
- Sector extensions
Check the extra conditions that apply in your sector.
- Healthcare extension
Check the extra conditions that apply in your sector.
- Education extension
Check the extra conditions that apply in your sector.
- Finance extension
Check the extra conditions that apply in your sector.
- Critical-infrastructure extension
Check the extra conditions that apply in your sector.
Worked examples
- Synthetic micro-business case — From shared inbox to validated copilot
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic SME case — A2 business agent for B2B quotes
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic nonprofit/foundation case — A2 agent for grant dossiers
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic public-sector case — A2 agent for planning dossiers
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic independent-professional case — From meeting to reviewed follow-up
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic independent-professional case — From copilot to an A2 business agent
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic independent-professional case — A3 orchestrated agency for one standard diagnostic
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic non-agentic case: Read-only RAG assistant for field procedures
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic non-agentic case: Weekly demand forecast for spare parts
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic non-agentic case: External customer-information chatbot
See a complete worked example. Its numbers are demonstrations, not results from your project.
- Synthetic non-agentic case: Multimodal product-catalogue accessibility assistant
See a complete worked example. Its numbers are demonstrations, not results from your project.
Templates
- AI accessibility assessment
Check whether people can understand and use the service, including with assistive tools.
- Evaluation plan
Choose test cases, acceptance criteria and a fair comparison.
- Anonymized field-feedback report
Record what was actually observed, including corrections and exceptions.
- Fundamental-rights impact assessment
Assess effects on people’s rights where this assessment is required.
- What to do during an AI incident
Prepare who stops the service, informs people and restores a safe situation.
- AI use-case mandate
Agree the task, owner, scope and stop rules before starting.
- Pilot decision
Record a decision to continue, change or stop, using actual test results.
- Initial risk assessment
Identify what could go wrong, who is affected and how to reduce it.
- AI literacy and training plan
Plan who needs to learn what, with practical exercises and checks.
- Use-case card
Describe the task, current process and expected output on one working card.
- AI supplier assessment
Compare suppliers on data handling, permissions, continuity and exit.
Methods and protocols
- AI use patterns
Separate the AI’s task from its human role, components and permissions.
- Tests and conditions for moving on
Define what to test and what must be true before moving on.
- First field-pilot cohort
Prepare a group of pilots with comparable records and review.
- Field-pilot protocol
Run a local comparison and count the human work that remains.
- Legal orientation — Switzerland and the European Union
Locate the Swiss and EU questions to check for your intended use.
- Maturity model
Identify what is already demonstrated and what still needs work.
- Local project dossier
Save, export and resume your project, then compare changes.
- Risk × autonomy classification
Choose an action boundary without confusing it with risk or capability.
- Security for AI and agentic systems
Choose security checks for the functions your system actually has.
- Task-time evidence and transfer
Estimate human hours saved or added from comparable task evidence.
- Universal AI adoption process
Follow the project from a defined task to operation and reassessment.
Evidence and references
- Evidence note: copilot, bounded automation, and strong automation
Check what a source measured and whether its result applies to your task.
- Public field-evidence review for AI adoption
Check what a source measured and whether its result applies to your task.
- Evidence note: AI-assisted independent knowledge work
Check what a source measured and whether its result applies to your task.
- Evidence note: AI in grant administration
Check what a source measured and whether its result applies to your task.
- Evidence note: B2B quoting in an SME
Check what a source measured and whether its result applies to your task.
- Reference note: how public administrations govern AI (2026)
Compare how administrations govern AI and build the profile for your jurisdiction.
- Evidence note: AI in public administration and planning
Check what a source measured and whether its result applies to your task.
- What recent implementations can offer you
Check what a source measured and whether its result applies to your task.
- Primary-source register
Check what a source measured and whether its result applies to your task.
- Evidence note: AI-assisted customer support in a small business
Check what a source measured and whether its result applies to your task.
Field evidence
- Field-feedback registry
Read or prepare a record of an observed pilot, separate from a simulation.