Hexa AI

Required License: Hexa AI is available by default with the Foundation and Advanced licenses. Tenable One customers with other licenses can request a free 90-day evaluation from their Tenable representative.

HexaAI is an AI-powered assistant integrated directly into the Tenable One Cloud Exposure Console. It provides a conversational interface to investigate, understand, and reason about your cloud environment using natural language. This eliminates the need to manually write complex queries, navigate deep menus, or pivot between different security tools.

Note: Hexa AI is not available in Gov Cloud environments.

Why Use Hexa AI?

Hexa AI helps you streamline security operations by:

  • Accelerating investigations — Get immediate answers in seconds instead of building manual filters or Explorer queries.
  • Iterative analysis — Follow security leads conversationally by asking follow-up questions to drill deeper into specific risks.
  • Simplifying complex data — Access technical data without needing to learn GraphQL schemas or specific query syntax.
  • Consolidating context — Hexa AI can simultaneously pull and correlate data from assets, findings, compliance frameworks, and official documentation.
  • Lowering barriers — A natural language interface makes the platform more accessible to new analysts or cross-functional team members.
Note: While Hexa AI makes suggestions, it requires a human in the loop to make decisions and perform actions. It cannot modify your cloud environment or Tenable configurations. For remediation, it identifies relevant findings and directs you to the appropriate console workflow.

Hexa AI Settings

Tenable Hexa AI settings are controlled via Tenable Vulnerability Management. To learn more, see the Vulnerability Management documentation.

Access Hexa AI

To launch the assistant, click the Hexa AI icon located in the top-right corner of the Cloud Exposure Console header.

The Hexa AI button in the Console header

To start a Hexa conversation, either type your prompt or click Dictate to use voice-to-text.

Add Files to Conversations

Hexa can work with files you upload or drag-and-drop directly in a conversation, so you don't need to reformat data by hand or describe a document you already have. Attach a spreadsheet, report, or diagram (e.g., an export of your resources and their owners, a diagram of your network topology) and ask Hexa to analyze, summarize, or extract information from it.

Files you upload stay available for the rest of that conversation and are not shared across conversations. Hexa can refer back to them in later messages, and you can download them from the chat at any time and delete them when you no longer need them.

Hexa supports the following file types: .csv, .gif, .html, .jpeg, .jpg, .md, .pdf, .png, .txt., .webp.

You can upload up to 10 files per conversation, and each file must not exceed 2 MB.

Core Abilities

Capability Description
Audit queries Retrieve audit trail and activity data via the GraphQL interface
Automation suggestions Suggest automations that send matching cloud or code findings to your teams, including the query and target scopes. You review and save the automation.
Cloud asset investigation Query and explore cloud resources (EC2 instances, S3 buckets, IAM roles, VMs, databases, and more) across cloud providers
Compliance reporting Summarize compliance posture against frameworks and generate structured reports
Documentation lookup Answer product how-to and configuration questions from official Tenable documentation
Exclusion suggestions Recommend and create policy exclusions for findings that are valid but non-actionable, including the scope, action, and parameters to apply
Identity & access review Investigate IAM principals, roles, permissions, and access paths
Policy and report building Construct Explorer-based queries to build custom policies or exportable reports
Relationship traversal Understand how entities relate to each other. For example, which workloads are exposed, or which identities have access to sensitive data.
Risk and findings analysis Retrieve and reason about security findings, policy violations, misconfigurations, and vulnerabilities.

Common Use Cases

Use Case Examples
Security investigation
  • Show me all S3 buckets that are publicly accessible and unencrypted.
  • Which EC2 instances have critical vulnerabilities and are internet-facing?
Identity and access
  • List IAM users with admin privileges who have not used their access in 90 days.
  • Which service accounts have overly permissive roles in GCP?
Compliance and reporting
  • Summarize our compliance posture against CIS AWS Foundations.
  • Generate a report of all high-severity findings in our production environment.
Cloud asset discovery
  • How many RDS instances do we have across all regions?
  • Show me all untagged resources in AWS us-east-1.
Policy building
  • Build a policy that detects publicly exposed databases with no encryption.
Exclusion management
  • Suggest an exclusion for findings on our dev, staging, and lab-tagged resources that keep showing up as non-actionable.
  • Recommend a severity downgrade instead of an exclusion for public load balancer findings in the QA account.
Automation setup
  • Create an automation for publicly exposed S3 buckets.
  • Create an automation for unencrypted RDS instances, but only for our production AWS accounts.
  • Create an automation for critical severity IaC findings.
How-to and product guidance
  • How do I onboard a new AWS account?
  • How does Tenable calculate effective permissions?

Advanced Use Cases

You can use Hexa AI to perform high-level analysis and generate complex artifacts by providing detailed, persona-based prompts. Below are examples of how to leverage Hexa AI for advanced security and operational tasks.

The following table provides advanced examples of how to leverage Hexa AI for complex security analysis and operational tasks.

Use Case Description Example
SLA Monitoring Analyzes open findings against strictly defined organizational resolution timelines to identify breaches and create remediation plans.

"Act as an IT Operations and Compliance Analyst. I am providing a dataset of open issues/tickets. My SLA (Service Level Agreement) policy is strictly defined as:

  • Priority: CRITICAL -> Resolve within 4 days.
  • Priority: HIGH -> Resolve within 7 days.

Please analyze the data and provide the following:

  1. SLA Breach Identification: Identify all findings that have currently exceeded their assigned SLA based on today's date ([Insert Today's Date]).
  2. Executive Summary: What are the most critical 'missed SLA' findings that pose the highest risk? Group these by age (e.g., 'Overdue by 30+ days').
  3. Immediate Action Plan: For the items identified, what are the next three tactical steps I should take to remediate these breaches immediately?
  4. Markdown Table: A detailed list of all breached items including: Ticket/Finding ID, Priority, Age (Days Open), Days Past SLA, Current Owner/Assignee, and Summary of Issue.

"

AI Workload Discovery Provides a comprehensive discovery plan to inventory AI software, machine learning libraries, and managed cloud AI services with sensitive data access.

"Act as a Cloud Exposure Architect and Data Governance Expert. I need a comprehensive discovery plan to locate and inventory all AI-related workloads across my cloud environment. The discovery must cover two specific criteria:

  1. AI Software & Packages (Application Layer): Identify Virtual Machines (EC2/GCE/Azure VM) and Container Images (ECR/GCR/ACR) that contain machine learning frameworks and libraries. Specifically scan for: pytorch, tensorflow, jupyter, scikit-learn, keras, pandas, transformers (HuggingFace), and open source tools like 'openclaw' or Ollama.
  2. Managed AI Services & Data Access (Infrastructure Layer): List all native cloud AI services in use and their associated IAM permissions/resource links. Check: SageMaker (AWS), Vertex AI (GCP), Azure Machine Learning, and Bedrock/OpenAI API integrations.

Critical Requirement: For each service found, identify if it has 'Allow' permissions to sensitive data buckets (S3/GCS/Blob) or databases (RDS/BigQuery) that are tagged as 'Confidential' or 'PII'.

Output results in a structured plan including a checklist for manual console audit and a summary table."

Toxic Combination Remediation Develops a multi-phased kill chain eradication plan addressing public exposure, critical vulnerabilities, and high-privilege roles.

"Act as a Senior Cloud Exposure Architect. I have an EC2 instance (ENTER NAME HERE) that is public-facing, contains critical OS-level vulnerabilities, and is attached to an IAM role with high privilege. Please provide a phased remediation plan to eradicate this kill chain. Include:

  • Immediate Containment — Steps to isolate the instance without breaking the application.
  • Privilege Reduction — A strategy to move from the current role to a Least Privilege policy.
  • Vulnerability Eradication — Options for patching vs. re-platforming.
  • Prevention — How to use AWS Service Control Policies (SCPs) to prevent this 'toxic combination' from occurring again.

"

Kill Chain Trade-off Analysis Analyzes complex risk vectors to determine and justify the single most impactful remediation action based on cost of attack versus ease of fix.

"Act as a Lead Cloud Exposure Architect. I have a critical risk on EC2 instance (ENTER NAME HERE). It is publicly accessible, has unpatched critical vulnerabilities, and carries excessive IAM privileges (AdministratorAccess). Instead of giving me a broad list of tasks, I want you to analyze the kill chain and decide on the single most impactful remediation action to take first. Please structure your response as follows:

  • The Decision — State clearly which of the three vectors (Network, Patching, or Identity) you would neutralize first and why.
  • The Logic — Compare the 'Cost of Attack' vs. 'Ease of Fix' for each vector to justify your choice.
  • The Implementation — Provide the specific technical steps (CLI or Console) to execute this 'single' fix.
  • Residual Risk — Briefly state what threat remains after this one fix is applied.

"

Operational Excellence Troubleshoots integration issues for cloud accounts and provides step-by-step plans to resolve "Connected with issue" statuses. "Investigate and resolve the integration issues associated with AWS Account ID (ENTER ACCOUNT HERE), explain why it is appearing in Tenable as Connected with issue. Provide a step-by-step remediation plan to fix the integration."
Cost Optimization Identifies "orphan" or unused AWS resources such as EBS volumes, Elastic IPs, and idle Load Balancers to reduce monthly cloud spend.

"Act as a Senior AWS Solutions Architect and FinOps Specialist. I need a solution to identify 'orphan' or unused resources in my AWS account that are currently incurring costs. Provide the following resource types in a Markdown table [Resource Type, Resource ID, Region, Reason, Estimated Cost]:

  1. Unattached EBS Volumes (status: available).
  2. Unassociated Elastic IP Addresses.
  3. Idle Elastic Load Balancers (ALB/NLB) with no targets.
  4. 'Stale' EBS Snapshots (older than 90 days and not associated with an AMI).
  5. Detached Elastic Network Interfaces (ENI).

For the 'Price' column, use standard AWS US-East-1 pricing estimates to provide a ballpark figure. Include a brief 'Cleanup' section with the AWS CLI commands needed to delete these resources safely."

Cloud Workload Health Monitors scan health by listing workloads with stale scans older than 24 hours to ensure continuous visibility and protection. "List all workloads where the 'Last Scan Date' is older than 24 hours from the current timestamp. Include the workload ID, name, and the exact time of the last successful scan. Sort the results by the oldest scan first."
Custom Data Type Definition Defines validation logic and regex patterns for proprietary or regional identification formats to support Data Security Posture Management.

"Task: Define a custom data type and validation logic for a Taiwan Identification Number (ROC National ID).

  • Format — Exactly 10 characters. Structure: [A-Z][1-2][0-9]{8}.
  • Character 1 — Uppercase letter (regional office).
  • Character 2 — Digit (1 for Male, 2 for Female).
  • Characters 3–9 — Seven random digits.
  • Character 10 — Checksum digit.

Validation Logic: Convert leading letter to code (A=10, B=11...). Assign weights: L1=1, L2=9, D1=8, D2=7, D3=6, D4=5, D5=4, D6=3, D7=2, D8=1, D9=1. Sum the products. Valid if S mod 10 = 0.

Output: Provide 5 valid examples and a Regex pattern for classification."