AI Feature Availability Guide
This guide explains how AI-powered features work across Deque's axe platform, what infrastructure is required to use them, and how administrators can configure them for their organization.
AI Features Overview
The Axe platform includes two categories of AI-powered features, each with different infrastructure dependencies.
AI Classification Features
These features are included as part of your axe DevTools Extension (Pro) / axe DevTools for Web subscription and do not consume AI credits:
- Role classification — AI-identification of element roles during Intelligent Guided Testing
- Table header detection — AI-identification of table headers
- Form label/input pairing — AI-assisted pairing of form labels and inputs
AI Credit-Consuming Features
The following features consume AI credits from your enterprise's monthly allocation and rely on Deque's LLM infrastructure:
| Feature | Product | What it does |
|---|---|---|
| Automated IGTs | axe DevTools Extension (Pro) / axe DevTools for Web | Fully automates Intelligent Guided Tests, allowing users to review and correct AI-generated results rather than manually stepping through each element |
| Advanced Rules | axe DevTools Extension (Pro) / axe DevTools for Web | Automated detection of accessibility issues that require visual or contextual reasoning, such as image alt text quality and text contrast |
| Remediate tool | axe MCP Server | Generates targeted, AI-powered remediation guidance for accessibility issues found during analysis |
The axe MCP Server analyze tool does not require LLM infrastructure and is available to all axe DevTools for Web customers regardless of AI configuration.
Infrastructure Requirements by Deployment Type
SaaS Customers
All AI features are available out of the box. No additional configuration is required.
Private Cloud Customers
Private cloud deployments require access to Deque's AI infrastructure components in order to use AI-powered features. There are two infrastructure components relevant to AI features:
- Deque ML Service — powers AI classification features (ML-suggested roles, table headers, form pairing)
- Deque LLM Service — powers AI credit-consuming features (Automated IGTs, Advanced Rules, MCP Server remediation)
Private cloud customers have two paths to enabling AI features, and these paths can be mixed and matched depending on your organization's requirements:
Option 1: Point to Deque's Shared AI Infrastructure
Your private cloud instance can be configured to connect to Deque's shared hosted AI services. This is the simplest path and requires no additional infrastructure on your end.
This option is available for the ML Service and LLM Service independently — you are not required to adopt both or neither.
Best for: Organizations that do not have strict requirements to keep all data processing within their own infrastructure.
Option 2: Private Instances of AI Infrastructure
Deque can work with your organization to stand up dedicated private instances of the AI infrastructure components. This keeps all data processing within your environment.
Best for: Organizations with strict data residency, compliance, or security requirements.
Standing up private instances of AI infrastructure components may involve additional cost. Please work with your Customer Success Manager for details.
On-Premises Customers
On-premises deployments requiring fully private AI infrastructure require escalation. Please contact your Customer Success Manager (CSM) to begin that process.
Feature Availability by Deployment Type
| Feature | SaaS | Private Cloud (shared AI infra) | Private Cloud (private AI infra) | On-Premises |
|---|---|---|---|---|
| AI classification features | ✅ | ✅ | ✅ | Contact your CSM |
| Automated IGTs | ✅ | ✅ | ✅ | Contact your CSM |
| Advanced Rules | ✅ | ✅ | ✅ | Contact your CSM |
MCP Server — analyze |
✅ | ✅ | ✅ | Contact your CSM |
MCP Server — remediate |
✅ | ✅ | ✅ | Contact your CSM |
Private cloud customers must be connected to the relevant AI infrastructure (shared or private) for a feature to be available. Features that depend on infrastructure your instance is not yet connected to will not be accessible until that connection is established. Work with your Customer Success Manager to get set up.
Administrator Configuration
Once AI infrastructure is connected and features are available for your instance, administrators can control AI behavior at the organization level via axe Configuration in the axe Account Portal.
AI-Powered Features
Found under: axe Configuration → Global Settings
This setting controls whether AI-powered features are enabled for your organization. When disabled, all AI capabilities are turned off across your entire organization, including:
- AI classification features (ML-suggested roles, table headers, form pairing)
- Automated IGTs
- Advanced Rules
- The axe MCP Server
remediatetool
Administrators can also choose whether individual users are permitted to override this setting within the axe DevTools Extension.
AI Data Sharing
Found under: axe Configuration → Global Settings
This setting controls whether testing data from your organization may be used to help improve Deque's AI models. This setting does not affect feature availability — it only governs data usage.
Administrators can also choose whether individual users are permitted to override this setting within the axe DevTools Extension.
Summary: Getting AI Features Enabled
If you are a private cloud customer and AI features are not yet available in your instance, the steps are:
- Determine your connectivity preference — shared Deque AI infrastructure, private instances, or a mix of both
- Work with your Customer Success Manager — they will coordinate establishing the appropriate infrastructure connections (note: private instances may involve additional cost)
- Verify your axe Configuration settings — once infrastructure is connected, confirm that "AI-Powered Features" is set to Enabled in your axe Configuration Global Settings
If you have questions at any step, reach out to your Customer Success Manager.