Diligent One agentic AI capabilities

Diligent One agentic AI capabilities provide a conversational experience that helps teams analyze information, identify risks, and support informed decisions. Powered by Claude Sonnet 4.6, an Anthropic large language model accessed through Amazon Bedrock, these capabilities deliver intelligent, real-time analysis so teams can proactively address issues before they affect delivery.

Diligent One agentic AI capabilities keep you in control at every step. You decide when to trigger AI analysis, review the insights it surfaces, and apply recommendations based on your professional judgment. This human-in-the-loop approach means AI enhances your expertise rather than replacing it.

Prerequisite

Toggle on the Enable AI features field to use the agentic AI capabilities. You can enable AI features from Projects, Risk Manager, or both, depending on the capabilities you want to use:

  • To use AI capabilities for audits, enable AI features from Projects > Settings > Enable AI features.

  • To use AI capabilities for risk management, enable AI features from Risk Manager > Settings > Global Settings > AI Permissions.

  • To use both sets of capabilities, enable AI features from both Projects and Risk Manager.

Enable AI features

When you enable AI features, the system uses Amazon Bedrock cross-region inference (CRIS) to power agentic AI capabilities using Claude Sonnet 4.6, an Anthropic large language model. CRIS gives you access to higher capacity and throughput than are available in your primary region alone.

When you enable AI features:

  • Your prompts and data may travel securely over AWS's private encrypted network to another AWS region within your geography for processing. The results return to you.

  • No data, logs, or other information is stored in the other region. Your data remains permanently stored only in your primary AWS region.

  • AI quality, accuracy, capacity, and throughput can improve.

  • Processing may occur in specific AWS regions depending on your primary region. The United Kingdom region is the exception: it uses a local in-region model instead—no cross-region processing occurs.

Where AI processing may occur

The AWS regions where processing may occur depend on your primary region:

Primary region

Processing details

Canada (ca-central-1)

US geographic profile (us.*). AI processing may occur in us-east-1, us-east-2, us-west-2

United States (us-east-1)

US geographic profile (us.*). AI processing may occur in us-east-1, us-east-2, us-west-2

European Union (eu-central-1)

EU geographic profile (eu.*). AI processing may occur in eu-west-1, eu-west-2, eu-west-3, eu-central-1, eu-north-1

United Kingdom (eu-west-2)

In-region model; no cross-region processing occurs

Australia (ap-southeast-2)

APAC geographic profile (au.* / apac.*). AI processing may occur in ap-southeast-1, ap-southeast-2, ap-northeast-1, ap-northeast-2, ap-south-1

Japan (ap-northeast-1)

APAC geographic profile (jp.* / apac.*). AI processing may occur in ap-northeast-1, ap-northeast-2, ap-southeast-1, ap-southeast-2, ap-south-1
South America

Global Cross - Region Interface (CRIS) routes AI inference requests to available AWS Regions for improved availability while keeping data at rest in the South America region.
Singapore Global Cross - Region Interface (CRIS) routes AI inference requests to available AWS Regions for improved availability while keeping data at rest in the Singapore region.

For additional details about cross-region inference, refer to Geographic cross-region inference in the Amazon Bedrock documentation.

If you disable AI features later, the features become unavailable in your region until you enable them again. For more information, contact your Diligent representative or go to Diligent customer support.

Use agentic AI capabilities

  1. Navigate to Task Manager.

  2. Select AI agent.

  3. Enter your question in the Ask Diligent One a question bar and select enter to send your request.

The system evaluates the following:

  • Task dependencies

  • Timeline constraints

  • Complexity and potential blockers

  • Risk description

  • Severity level

  • Likelihood or impact

Review the analysis, apply recommendations as appropriate, and repeat the cycle as tasks evolve. The conversational experience retains context within a session, so you can refine results through follow-up prompts.

Capabilities

Agentic AI capabilities currently support the following:

Project setup and engagement scoping

  • Set up projects and define engagement scope.

  • Draft clear and structured audit objectives.

Risk identification and assessment

  • Automatically surface potential risks for engagements.

  • Create and configure risk assessments.

Risk-to-control mapping

  • Map identified risks to relevant controls.

  • Ensure proper mitigation coverage and traceability.

Related resources