Stay organized with collections
Save and categorize content based on your preferences.
What is AML AI?
AML AI is an API-based machine learning pipeline for
automatically training, testing, deploying, and monitoring a productionized
anti-money laundering (AML) model. As a managed service, Google takes care of
the infrastructure behind the scenes and presents teams with a production-ready
system to train, predict, and backtest models to tackle money laundering.
Interface
The main way of interacting with AML AI API is using the
https://financialservices.googleapis.com endpoint with REST API calls. The
Google Cloud CLI tool is not supported for directly calling the
AML AI API, but it is recommended to use the Google Cloud CLI to
obtain credentials.
You may want to use programming languages to interact with AML AI.
To make coding against AML AI easier, Google provides
generic API client libraries
for a number of different languages that can reduce the amount of code you need
to write and make your code more robust.
Each of the API client libraries provide a means to use application default
credentials (ADC).
AML AI reads input data from BigQuery and writes output
predictions and backtest data to BigQuery. For input data, an
AML AI dataset resource must be created which references the
data in BigQuery. This dataset must reside in the same location as the
AML AI instance.
The AML AI dataset resource represents pointers to datasets 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). Modifying the data this way is not recommended. For more
information, see
Create and manage datasets.
Services used by AML AI
As well as the AML AI API itself, there are a number of other
Google Cloud API services which are required to use the AML AI:
Required
Cloud IAM: for identity management and access management
Cloud KMS: for key management
BigQuery: for data storage
Cloud Logging: for logging and monitoring
Optional
Cloud HSM: Optional hardware-backed storage for encryption keys
VPC Service Controls: Prevent data exfiltration to unauthorized networks and devices
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-08-29 UTC."],[[["\u003cp\u003eAML AI is a managed, API-based machine learning pipeline that automates the training, testing, deployment, and monitoring of anti-money laundering models.\u003c/p\u003e\n"],["\u003cp\u003eInteraction with the AML AI API is primarily through the \u003ccode\u003ehttps://financialservices.googleapis.com\u003c/code\u003e endpoint using REST API calls, and while the Google Cloud CLI is not supported for direct calls, it is recommended for obtaining credentials.\u003c/p\u003e\n"],["\u003cp\u003eAML AI utilizes BigQuery for both reading input data and writing output predictions and backtest data, requiring the creation of an AML AI dataset resource that references the relevant data in BigQuery.\u003c/p\u003e\n"],["\u003cp\u003eSeveral Google Cloud services are essential for AML AI, including Cloud IAM, Cloud KMS, BigQuery, and Cloud Logging, while Cloud HSM and VPC Service Controls are optional for enhanced security.\u003c/p\u003e\n"],["\u003cp\u003eWhile no API-specific Cloud client libraries exist for AML AI, generic client libraries are available to reduce coding needs and enhance code robustness when interacting with AML AI.\u003c/p\u003e\n"]]],[],null,["# Architectural overview\n\nWhat is AML AI?\n---------------\n\nAML AI is an API-based machine learning pipeline for\nautomatically training, testing, deploying, and monitoring a productionized\nanti-money laundering (AML) model. As a managed service, Google takes care of\nthe infrastructure behind the scenes and presents teams with a production-ready\nsystem to train, predict, and backtest models to tackle money laundering.\n\nInterface\n---------\n\nThe main way of interacting with AML AI API is using the\nhttps://financialservices.googleapis.com endpoint with REST API calls. The\n[Google Cloud CLI tool](/cli) is not supported for directly calling the\nAML AI API, but it is recommended to use the Google Cloud CLI to\nobtain credentials.\n\nYou may want to use programming languages to interact with AML AI.\nTo make coding against AML AI easier, Google provides\n[generic API client libraries](/apis/docs/client-libraries-explained#google-api-client-libraries)\nfor a number of different languages that can reduce the amount of code you need\nto write and make your code more robust.\n\nEach of the API client libraries provide a means to use application default\ncredentials (ADC).\n| **Note:** There are no API-specific Cloud client libraries for AML AI.\n\nFor details about the REST interface, see\n[Financial Services API](/financial-services/anti-money-laundering/docs/reference/rest).\n\nData\n----\n\nAML AI reads input data from BigQuery and writes output\npredictions and backtest data to BigQuery. For input data, an\nAML AI dataset resource must be created which references the\ndata in BigQuery. This dataset must reside in the same location as the\nAML AI instance.\n\nThe AML AI dataset resource represents pointers to datasets in\nBigQuery. It does not hold or point to any specific snapshot of the\ndata in these tables. If data is modified after a dataset is created (for\nexample, if records are deleted), this will be reflected in the results of other\ncalls to the API (for example, the creation of new models or when running\npredictions). Modifying the data this way is not recommended. For more\ninformation, see\n[Create and manage datasets](/financial-services/anti-money-laundering/docs/create-and-manage-datasets).\n\nServices used by AML AI\n-----------------------\n\nAs well as the AML AI API itself, there are a number of other\nGoogle Cloud API services which are required to use the AML AI:\n\nRequired\n\n- Cloud IAM: for identity management and access management\n- Cloud KMS: for key management\n- BigQuery: for data storage\n- Cloud Logging: for logging and monitoring\n\nOptional\n\n- Cloud HSM: Optional hardware-backed storage for encryption keys\n- VPC Service Controls: Prevent data exfiltration to unauthorized networks and devices\n\nWhat's Next\n-----------\n\n- [Security and compliance features](/financial-services/anti-money-laundering/docs/concepts/security-and-compliance-features)"]]
