Intended use of the AI system

Make sure the AI system is used in line with its intended uses and its accompanying documentation. In practice the record is deployment evidence following system instructions and documentation. For intended use of the ai system the evidence is what people did, not what they were told.

What it asks

Make sure the AI system is used in line with its intended uses and its accompanying documentation.

Where generative AI changes the answer

Intended use has to exclude the uses the system was not built for, and those exclusions have to reach the people who could attempt them.

Who it binds

The business owner of the use and the people who act on the output. For A.9.4 the named role is the owner of deployment evidence following system instructions and documentation.

Evidence an auditor normally samples

• Deployment evidence following system instructions and documentation
• Checks that input data match documented conditions
• Event logs demonstrating use as intended
• Records of concerns raised internally and to suppliers

The test you can run

Compare deployment evidence following system instructions and documentation with the use as it runs today; the gap between the two is the finding to record.

Common failures

• System used on data outside its documented domain
• No logs to show use as intended

Before the draft publishes

For A.9.4, write the acceptable-use rule in the words staff use, publish it where they work, and record the first exception it produces. The first artefact to produce is deployment evidence following system instructions and documentation.

Where this sits against the draft

The draft's published scope names controls for the risks it identifies, and A.9.4 (intended use of the ai system) is the published ISO/IEC 42001 control closest to that work. The draft does not replace it, and an audit still tests it against the evidence listed here.