Safe AI Use on SAP Projects: Protect Client Data

SAP systems hold a company's most sensitive data: finances, pricing, suppliers, customers, employees and production plans. As a consultant, you are usually bound by a contract and a confidentiality agreement with the client. Using AI carelessly can break those agreements and damage trust that took years to build. This lesson turns the course into a safe daily workflow.

Know your obligations first

Before using any AI tool on a project, find out:

Your employer's AI policy (consulting firm or company)
The client's AI policy and any clauses in the contract or NDA about third-party tools
Which AI tools are approved for project work, and which data they may be used with

If nobody can tell you, ask your project manager in writing. "I didn't know" is not a good position on a client project.

Never paste into an unapproved tool

Client name together with internal project details
Real master data: vendors, customers, materials with prices, bank details
Employee data, including anything from HR or payroll
Financial figures, pricing conditions, margins
System access details: user IDs, passwords, connection details, system IDs and hostnames
Full custom code or specifications that the client considers confidential
Screenshots that show any of the above

Anonymise by default

Real	Replace with
Client and group company names	Client A, Company B
Company codes, plants, sales orgs	1000, 2000 (neutral examples)
Vendor / customer names and numbers	Vendor X, Customer 1
Amounts and prices	Rounded or invented example values
User names	User A
Custom object names that reveal the client	ZOBJECT_1

Most questions (how to structure a spec, why an error happens, how to write a test) can be answered with anonymised examples just as well.

A daily safe-use workflow

Classify the task: Green (drafting), Amber (assisted analysis) or Red (do not use AI), from lesson 1.
Use the approved tool for that kind of work.
Anonymise before pasting.
Verify every system-specific statement in the system, in SAP documentation or SAP Notes.
Review like you would review a junior colleague's work before it reaches the client.
Record important AI use where your project requires it, for example in working notes.

Being transparent with the client

Many clients are open to AI-assisted work if it is done transparently and safely. A short, honest statement in your approach can build trust:

Draft a short paragraph for our project approach document explaining how
the team uses AI tools: only approved tools, no client confidential data
in public tools, all outputs reviewed by consultants, and SAP-specific
content verified in the system. Professional tone, under 120 words.

Agree the wording with your project manager before sharing it.

If something goes wrong

If you realise confidential client data went into a tool that was not approved:

Stop and do not paste anything more.
Delete the conversation if the tool allows it.
Inform your project manager or the person responsible for data protection, following your firm's process.

Fast, honest reporting limits the damage. Hiding it makes it worse.

Final practice task

Write a one-page "AI on this project" note for yourself or your team: approved tools, Green/Amber/Red tasks, the anonymisation table, verification sources (system, SAP Help Portal, SAP Notes), and who to contact if something goes wrong. Share it with your project manager.

Then mark this lesson complete. You have finished the course!

Where to go next

Improve every prompt with Prompt Writing That Works.
For ABAP, debugging and testing in more depth, see AI for Developers.

Key takeaways

Know your employer's and the client's AI policies before using any tool.
Never paste client master data, financials, employee data or access details into unapproved tools.
Anonymise by default, verify in the system, and review before anything reaches the client.
Be transparent about AI use, and report mistakes quickly.