Exfiltration: AI Agent Initiated BigQuery Data Exfiltration to External Table

Preview

This feature is subject to the "Pre-GA Offerings Terms" in the General Service Terms section of the Service Specific Terms . Pre-GA features are available "as is" and might have limited support. For more information, see the launch stage descriptions .

This document describes a threat finding type in Security Command Center. Threat findings are generated by threat detectors when they detect a potential threat in your cloud resources. For a full list of available threat findings, see Threat findings index .

Overview

A resource was saved outside of your organization or project initiated by an AI agent.

Event Threat Detection is the source of this finding.

How to respond

To respond to this finding, do the following:

Step 1: Review finding details

1. Open the Exfiltration: AI Agent Initiated BigQuery Data Exfiltration to External Table finding, as directed in Reviewing findings .

2. On the Summarytab of the finding details panel, review the listed values in the following sections:

   • What was detected:
     • Principal email: the account used to exfiltrate the data.
     • Exfiltration sources: details about the tables from which data was exfiltrated.
     • Exfiltration targets: details about the tables where exfiltrated data was stored.
   • Affected resource:
     • Resource full name: the full resource name of the project, folder, or organization from which data was exfiltrated.
   • Related links:
     • Cloud Logging URI: link to Logging entries.
     • MITRE ATT&CK method: link to the MITRE ATT&CK documentation.
     • Related findings: links to any related findings.

3. Click the Source Propertiestab and review the fields shown, especially:

   • evidence :
     • sourceLogId :
       • projectId : the Google Cloud project that contains the source BigQuery dataset.
   • properties
     • dataExfiltrationAttempt
       • jobLink : the link to the BigQuery job that exfiltrated data.
       • query : the SQL query run on the BigQuery dataset.

4. Optionally, click the JSONtab for the complete listing of the JSON properties of the finding.

Step 2: Review permissions and settings

1. In the Google Cloud console, go to the IAMpage.

   Go to IAM

2. If necessary, select the project listed in the projectId field in the finding JSON.

3. On the page that appears, in the Filterbox, enter the email address listed in Principal emailand check what permissions are assigned to the account.

Step 3: Check logs

1. On the Summary tabof the finding details panel, click the Cloud Logging URIlink to open the Logs Explorer.

2. Find admin activity logs related to BigQuery jobs by using the following filters:

   • protoPayload.methodName="Jobservice.insert"
   • protoPayload.methodName="google.cloud.bigquery.v2.JobService.InsertJob"

Step 4: Research attack and response methods

1. Review the MITRE ATT&CK framework entry for this finding type: Exfiltration to Cloud Storage .
2. Review related findings by clicking the link on the Related findingson the Related findingsrow in the Summarytab of the finding details. Related findings are the same finding type on the same instance and network.
3. To develop a response plan, combine your investigation results with MITRE research.

Step 5: Implement your response

The following response plan might be appropriate for this finding, but might also impact operations. Carefully evaluate the information you gather in your investigation to determine the best way to resolve findings.

• Contact the owner of the project with exfiltrated data.
• Consider revoking permissions for userEmail until the investigation is completed.
• To stop further exfiltration, add restrictive IAM policies to the impacted BigQuery datasets ( exfiltration.sources and exfiltration.targets ).
• To scan impacted datasets for sensitive information, use Sensitive Data Protection . You can also send Sensitive Data Protection data to Security Command Center . Depending on the quantity of information, Sensitive Data Protection costs can be significant. Follow best practices for keeping Sensitive Data Protection costs under control .
• To limit access to the BigQuery API, use VPC Service Controls .
• To identify and fix overly permissive roles, use IAM Recommender .

What's next

• Learn how to work with threat findings in Security Command Center .
• Refer to the Threat findings index .
• Learn how to review a finding through the Google Cloud console.
• Learn about the services that generate threat findings .