Aigcchecker Product Update: Writing Feedback Beta, Mixed Classification, and Platform Improvements

Aigcchecker continues to build in public with the goal of evolving alongside its users. To support this approach, the team has launched a bi-weekly product update that consolidates newly released features, enhancements, bug fixes, and early previews of what is coming next. This update covers the work completed over the last two weeks and highlights meaningful progress across writing assistance, AI detection transparency, and platform stability.

The focus of this release is twofold: improving how users understand AI detection results and expanding Aigcchecker’s role as a writing support tool rather than a purely evaluative system.

Writing Feedback Is Now in Beta

The most notable addition in this update is the introduction of Writing Feedback, a new feature designed to help users actively improve their writing. Unlike detection-focused tools, Writing Feedback provides actionable suggestions across multiple dimensions, including document structure, clarity, conciseness, grammar, spelling, and overall fluency.

The feature is currently available in beta to a subset of users. Performance optimization is ongoing, with real-time streaming of feedback results in development to significantly reduce wait times and improve the overall experience.

Dashboard Features and Enhancements

The Aigcchecker dashboard has been updated to present detection results with greater nuance and clarity. Results now include a mixed classification category, in addition to AI and human classifications. This is paired with confidence scores that help users better interpret borderline or blended cases.

A new visual representation displays mixed results as a sphere indicating the proportion of AI involvement. This update also resolves a prior issue where deep scans with mixed predictions failed to highlight specific sentences.

Additional Dashboard Improvements

To support multi-class detection, batch upload and scan history table fields have been renamed from “AI Probability” to the more inclusive “Probability.” Plagiarism scanning is now supported for batch file uploads, making large-scale workflows more efficient.

Scan configuration toggles for Writing Feedback, Deep Analysis, and Plagiarism are now persisted. These selections remain active across scans and browser refreshes, reducing repetitive setup and improving usability.

Bug Fixes

Several issues affecting accuracy and user experience have been addressed. Highlighting problems in deep scans with mixed predictions have been fixed, ensuring sentence-level insights are consistently displayed.

An issue that caused free personal plan users and paid team members to be incorrectly limited when team usage exceeded 10,000 words has also been resolved.

Team Feature Updates

Team functionality has been expanded with the ability to invite or import members via file upload. Additional stability improvements have been implemented to ensure smoother collaboration and more reliable usage tracking across teams.

Main Website Improvements

The main Aigcchecker website now reflects the same mixed classification and confidence score experience available in the dashboard, including the new visual indicator for mixed results. Plagiarism scanning and AI copyright checks have also been enabled on the main site, expanding access to these capabilities.

Frequently Asked Questions (FAQ)

What is Writing Feedback in Aigcchecker?

Writing Feedback is a new beta feature that provides suggestions on structure, clarity, grammar, spelling, conciseness, and fluency. It is designed to help users improve their writing, not just evaluate whether it was AI-generated.

Who can access the Writing Feedback beta?

Writing Feedback is currently available to a limited group of users. Aigcchecker is gradually expanding access while improving performance and preparing the feature for a wider release.

What does “mixed classification” mean?

Mixed classification indicates that a document likely contains both human-written and AI-assisted content. It provides a more nuanced alternative to binary AI or human labels.

How are confidence scores used?

Confidence scores show how certain the system is about its classification. They help users interpret results more responsibly, especially in edge cases where content characteristics overlap.

Does Aigcchecker store or reuse uploaded content?

Aigcchecker is designed with privacy considerations in mind. Users should review the platform’s privacy policy for details on data handling, storage, and compliance with relevant regulations.

Looking Ahead

This bi-weekly update reflects Aigcchecker’s ongoing effort to improve transparency, usability, and writing quality support. Future updates will continue to document feature rollouts, refinements to detection accuracy, and new tools currently in development.