AIGC-X
The AI-generated content detection tool jointly launched by the State Key Laboratory of People's Network, University of Science and Technology of China, and the Institute of Artificial Intelligence at Hefei Comprehensive National Science Center supports text and image detection and provides confidence levels.
Updated: 2026-10-26
What is AIGC-X?
AIGC-X is an AI-generated content detection tool jointly launched by the National Key Laboratory of Content Cognition and Communication at People's Daily, University of Science and Technology of China, and the Hefei Comprehensive National Science Center Institute of Artificial Intelligence. According to official descriptions, it can quickly distinguish whether content is machine-generated or human-generated, and also provides two detection entrances: AI-generated text and AI-generated images. The official states that it focuses on the detection of general genres.
Who is it for?
- Media editors and fact-checkers: Initially screen whether manuscripts are machine-generated to provide clues for subsequent verification.
- Teachers and academic administrators: Identify clues about the originality of texts submitted by students and use them as a basis for communication.
- Content platform review teams: Conduct pre-risk screening for UGC content to reduce the pressure on manual review.
- Researchers: Refer to its detection conclusions and confidence levels to conduct research on AI-generated content and detection technologies.
- Corporate Communications Roles: Verify the authenticity of externally contributed content and avoid reposting unsubstantiated AI-generated content.
Key features
6AI Text Detection
The official states that it captures implicit features such as perplexity and emergent frequency through deep language modeling to determine whether the text is machine-generated.
AI Image Detection
AIGC-XP captures implicit features such as fingerprints of typical generative models and traces in small areas through deep visual modeling to identify generated images.
Algorithm Fusion Framework
The official states that it adopts an AI framework integrating algorithm fusion and knowledge-driven approaches, combined with content recognition and content traceability technologies.
Confidence Score Output
The detection results will provide a judgment on whether the content is machine-generated or human-generated along with the corresponding confidence score value, facilitating hierarchical processing.
Publicly Stated Limitations
The official actively notes that the accuracy for short texts is limited, it currently only supports Chinese, and its robustness to specific genres and adversarial modifications is weak.
For AI Governance
The official states that this tool is an important achievement in researching AI self-governance mechanisms and promoting AI risk management and safety review and evaluation.
How to use AIGC-X
- The official explicitly notes that the detection accuracy is limited for short and concise texts, and conclusions should not be drawn based on the detection results of just a few sentences.
- Currently, it only supports Chinese content detection, is not applicable to foreign language texts, and the results of mixed text will also be affected.
- The official explanation states that detection results will shift due to manual modification, and artificially rewritten content is more likely to be judged as human-generated.
- The test conclusions are for reference only and do not constitute the review and confirmation of the platform or institution.
- important judgments require manual review.
- Generative technology is constantly evolving, and detection is inherently adversarial.
- It is recommended to use it only after cross-verification with other evidence.
- In addition, it should be specially noted that AI detection outputs a "probability judgment" rather than a final conclusion: misjudgments are not uncommon in scenarios such as short texts, content dense with professional terminology, or texts that have been deeply manually rewritten.
- It is not recommended to use it as the sole basis for academic misconduct or content violations.
Traffic & popularity
Data month Sep 2026
AIGC-X receives about 601 visits per month.
Monthly visits
12 monthsTraffic sources
- Direct66.6%
- Referrals27.1%
- Search6.2%
Top countries
- China100.0%
Estimated from third-party traffic data.
Similar tools
DayaThe document detection system under Chaoxing provides similarity detection, AIGC detection, format detection, one-click typesetting and AI intelligent evaluation. The official claims that it relies on book resources to achieve dual detection of books and periodicals.
Zhuque AI DetectionTencent has launched a free AI-generated content detection tool. Trained on a variety of advanced models and millions of data points, it delivers particularly outstanding performance in Chinese detection. It supports both text and image detection and provides an API for easy integration into review processes.
CNKI AIGC DetectionCNKI's plagiarism check and AIGC detection services for individual users adopt the adaptive multi-order semantic fingerprint detection technology, which discriminates the AI generation features of text through two links: language patterns and semantic logic, according to the official statement.
Proofig AIProofig AI is an image integrity review tool for scientific research publishing. It detects image tampering, repeated reuse, and AI-generated images. The official statement says it is used for self-checks before submission to reduce the risks of rejection and retraction.
WenchaoThe academic quality inspection platform launched by Wanfang Data can simultaneously complete multi-element detections such as text similarity, text AIGC, image reuse, reference risk and language specification in one submission.
Weipu AIGC DetectionThe Weipu Paper Detection System provides plagiarism check, format checking, AIGC detection and intelligent evaluation services, which are distinguished by versions for college students, postgraduates, editorial departments and professional titles, and supports unified management of personal versions and institutional versions.