Note: This article is updated weekly throughout the month as new SafeBase features are released. All updates are also published on our Product Updates site.
Table of Contents
Multi-Select Answers in AIQA
AI Questionnaire Assistance (AIQA) now supports selecting multiple answers for a single question, so teams can accurately respond to “select all that apply” questions without workarounds.
What’s New
Select more than one valid answer option for any multi-select question in AIQA.
Single-answer questions continue to behave exactly as before.
Multi-select support is available anywhere answers are selected in AIQA.
Evidence Library Sync
Evidence Library Sync bridges the gap between internal compliance operations in Drata and external security assurance in the Trust Center. With this enhancement, GRC teams can automatically sync selected compliance evidence from Drata into their Trust Center or AI Questionnaire Automation environment. This eliminates the need for manual uploads, reduces lag in security reviews, and ensures external stakeholders always see the most up-to-date artifacts—saving time and accelerating trust.
Learn more about Syncing Evidence from Drata to the SafeBase Trust Library.
Knowledge Base Update: AI-Powered Potential Issues
What is the Potential Issues Feature?
We have introduced an automated detection and remediation system for Knowledge Base (KB) maintenance. This feature uses AI to identify outdated, duplicate, or conflicting entries that reduces the quality of AI Questionnaire Assistance and Trust Center search results.
The Benefit
Historically, maintaining a Knowledge Base required exhaustive manual audits. By transitioning to an “AI-detected, human-verified” workflow, customers significantly reduce overhead while ensuring that all downstream AI-generated responses remain accurate, consistent, and professional.
Solution Highlights
Automated Issue Detection: Scans the KB to flag invalid Q&As, unclear answers, and contradictory entries.
Intelligent Bulk Cleanup: Enables rapid resolution of flagged issues. Users can edit or delete entries en masse after a required manual review.
Granular Control: No deletions occur without explicit user confirmation, ensuring "human-in-the-loop" oversight for data integrity.
Technical Callouts
GenAI Requirement: Generative AI must be enabled for full detection capabilities. Without it, the system provides duplicate and similar question detection via traditional matching.
Configuration: Proper setup is essential. Customers should establish their auto-resolve preferences (Knowledge Base > Potential Issues > Configuration) before they start using this feature. Once enabled, customers will be prompted to configure these settings the first time they navigate to the Potential Issues page.