This page shows you how to resolve issues with AML AI.

## Permission errors

This section lists permission errors that you might encounter and provides
suggestions for how to fix each of them.

### The service account doesn't have permission to access the KMS key

The API automatically creates a service account in your project. The service
account needs access to the CMEK key so it can use the key to encrypt and
decrypt the underlying data. (It also needs read access to the BigQuery input
dataset to be used for creating models, and it needs write access to the
BigQuery output dataset.) The following error occurs when the service account
does not have permission to access the KMS key:

```
Error code 400 - service-account@gcp-sa-financialservices.iam.gserviceaccount.com
lacks permissions to KMS key
```

To resolve this issue, grant access to the key. For more information, see
[Grant access to the CMEK key](https://docs.cloud.google.com/financial-services/anti-money-laundering/docs/set-up-instance.md.txt#grant-access-cmek).

## Dataset issues

This section lists dataset issues that you might encounter and provides
suggestions for how to fix each of them.

### Unexpected results

A Financial Services API dataset resource is a pointer to data in BigQuery. It
does not hold or point to any specific snapshot of the data in these tables. If
data is modified after a dataset is created (for example, if records are
deleted), this will be reflected in the results of other calls to the API (for
example, the creation of new models or when running predictions).

To reduce the occurrences of unexpected results, be sure to maintain the
integrity of the data in BigQuery.