Manage AI Features on Your Appian Environment

This page describes how to control which AI features are active on your Appian environment, including how to disable AI capabilities and manage new AI functionality as it becomes available.

Default state of AI features by deployment type

The default state of AI features depends on your deployment type:

• Appian Cloud: When your environment is provisioned with a license tier that includes access to AI features, those AI features are enabled by default.
• Self-managed: AI features are disabled by default. Customers must submit a support ticket to request that AI features be enabled on their environment. Some AI features may not be available, or may have limited functionality, in self-managed deployments of Appian. For a list of supported capabilities, see Set Up Appian AI for Appian on Kubernetes.

For a complete list of Appian's AI capabilities on all deployments, see Appian's AI capabilities.

Disable AI features

All AI features can be disabled except for the AI Skill design object. However, you can request to disable generative skill types within the AI Skill design object.

To request that AI features be disabled on your environment, submit a support ticket through Appian Support specifying which features you want disabled.

You can request to disable any individual AI feature, or all AI features at once, from the set of AI capabilities that support disablement.

Enable new AI features after previously opting out

Customers who have disabled AI features maintain full control over when and which AI capabilities become active on their environment. If you have previously requested that AI features be disabled on your environment, Appian applies the following rules when new AI features are released:

• New AI features are not automatically enabled. Newly released AI functionality is not activated on environments where AI features have been previously disabled. This rule applies to all AI features, whether or not the new functionality is related to the features you previously disabled.
• Opting in to new AI features requires a support request. If you want to enable a newly released AI feature, you must submit a support ticket requesting that the specific functionality be enabled on your environment.

Disable generative AI skill types

The AI skills framework includes both traditional machine learning (ML) skill types and generative AI skill types. The AI Skill design object cannot be fully disabled at the environment level. Because many organizations have stricter governance requirements around generative AI than traditional ML, Appian allows you to disable specific AI skill types that use generative AI while keeping ML-based AI skill types available.

To disable generative AI skill types on your environment, submit a support ticket through Appian Support specifying which generative AI skill types you want disabled.

Disable a model provider in the Admin Console

You can disable individual model providers directly from the Admin Console. Disabling a model provider prevents Appian from routing any AI requests to that provider's models. It also removes those models from the fallback chain and records the change in the audit log.

Before disabling a provider, verify that no active applications or features depend on that provider's models. For details on how providers apply to other AI features, see AI Services.

To disable a model provider:

1. In the Appian Admin Console, go to AI Services.
2. Click SETTINGS.
3. Locate the provider you want to disable, such as Anthropic, Amazon, Meta, or OpenAI.
4. Toggle the provider from Enabled to Disabled.
5. Review the confirmation message describing the impact on AI capabilities.
6. Click SAVE CHANGES.

Feedback