Release Date: 2026-03-11
Major enhancement to the Essay question type, including AI-powered scoring capabilities, improved rubric-driven evaluation, and a streamlined scoring workflow.
These enhancements are designed to scale open-ended assessments, improve grading consistency, and reduce manual effort while maintaining control, transparency, and defensibility in scoring.
New Features
1. AI-Powered Essay Scoring (AI Services)
Introduced AI Services that enable automated and assisted grading for Essay responses.
Scoring Options
Organizations can now choose how Essay responses are graded:
- Manual Scoring (Default)
- No AI involvement
- All responses require human evaluation
- AI-Assisted Grading
- AI generates an initial score
- A human reviewer must review, confirm, or adjust the score
- Reviewers can provide feedback and context before finalizing
- Fully Automated AI Grading
- AI automatically assigns scores without requiring human approval
- Reviewers can still modify scores if needed and add justification
Key AI Capabilities
- Rubric-aligned scoring ensuring results remain consistent with predefined evaluation criteria
- Plagiarism / AI-generated text detection to flag potentially inauthentic responses
- Prompt injection detection to identify attempts to manipulate AI scoring
2. AI Services Enablement Framework
AI scoring is controlled through a structured enablement process:
3. Enhanced Essay Question Structure
Essay questions have been redesigned to support more advanced scoring models:
4. Improved Essay Scoring Workflow
The scoring experience has been redesigned to improve efficiency:
New Filtering Options
- Need Review – pending responses
- Recently Graded – recently completed evaluations
Enhancements to Existing Functionality
Rubric Improvements
- Rubrics are now directly embedded into the scoring interface
- Existing rubric structures are automatically converted to the new model
- Backward compatibility is preserved (default 100% weight applied to legacy rubric)
Essay Scoring Experience
Building on existing functionality:
Flexible Adoption of AI
- Organizations retain full control over:
- Whether AI is used
- How AI is used (assist vs. automate)
- When human oversight is required
Security and Integrity Controls
Included safeguards to ensure responsible AI usage:
Compatibility & Migration
- Existing Essay questions remain fully functional
- Legacy rubrics are automatically converted to the new weighted model
- No customer action required to preserve current scoring behavior
Comments
0 comments
Article is closed for comments.