Version: 2026-06-05
Summary
- Explain AI use calmly; do not hide it and do not exaggerate it.
- Clarify before project start whether and how AI may be used.
- If AI is fully excluded, that can affect effort, speed, and cost.
- Sensitive data, secrets, and production data are taboo without explicit approval.
- Promise only what is technically and organisationally controlled.
- No AI must be defined concretely: cloud, local, code, documentation, review, project management, external contributors.
- When something is suspected, clarify facts first and then communicate in a coordinated way.
Topic
Customer-facing communication about AI use in software development projects.
Starting Point
AI is now a normal topic in development. Some customers expect it, some worry about it, and some forbid it until rules are clear.
A calm explanation is better than hype or concealment. The customer does not need tool marketing. They need to understand what is used, what is excluded, and who remains responsible.
Risk
Vague statements create trouble. We use AI responsibly is not enough. No AI is also unclear unless it defines whether it covers cloud tools, local tools, generated code, documentation, reviews, project management, or external contributors.
Recommendation
Communicate AI use as a controlled development aid. Be concrete about boundaries: no secrets, no production data, no customer data without approval, human review, project-specific exclusions, and clear responsibility.
If a customer forbids AI completely, document the scope and consequences. Full exclusion may reduce speed or require different working methods.
Review Checklist
- Has AI use been discussed before project work starts?
- Are customer data and secrets explicitly excluded?
- Are local and cloud tools distinguished?
- Is responsibility clearly human?
- Are external contributors covered?
- Are incident communications coordinated and fact-based?
Wall rules
Explain AI use calmly, do not hide it.
Only promise what we control.
No customer data in AI without approval.
No AI must be defined concretely.
Clarify facts first, then inform customers.
Open questions for a project
- Which AI tools may be used for this customer?
- Which data categories are excluded?
- How should AI use be described in project communication?
- Who speaks to the customer in case of an incident?