Overview
AIGCTrustDetect hosts a three-track shared task benchmark on trustworthy AIGC detection and governance. The tracks move beyond binary labels toward fine-grained, cross-lingual and content-level evaluation.
The organizing team previously ran the shared tasks on LLM-generated text detection at NLPCC 2025 (Task 1) and NLPCC 2026 (Task 6), which together attracted more than 120 participating teams and resulted in six accepted system description papers, as well as the NLPCC 2026 shared task on the reliability of AI-assisted scientific reports (Task 10).
Want a feel for each track first? The Home page has one interactive example per track.
Track 1. Multilingual LLM-Generated Text Detection
This track covers eight languages and performs fine-grained classification across human-written, LLM-generated and LLM-refined texts. It evaluates the robustness and cross-lingual generalization of detection methods under real-world deployment scenarios.
| Task | Three-way classification: human-written, LLM-generated, LLM-refined |
|---|---|
| Languages | Eight languages [list TBD] |
| Data | [TBD] |
| Shared task homepage | [link TBD] |
| Organizers | [TBD] |
| Registration | [link TBD] |
Track 2. Multilingual LLM-Generated Archaic Text Detection
This track targets AI-generated texts written in classical and archaic languages, filling a gap in content governance for historical, literary and classical-style corpora.
| Task | Detection of AI-generated text in classical and archaic languages [label set TBD] |
|---|---|
| Languages | [TBD] |
| Data | [TBD] |
| Shared task homepage | [link TBD] |
| Organizers | [TBD] |
| Registration | [link TBD] |
Track 3. LLM-Generated Text Reliability Assessment
This track focuses on overstated claims and evidence–statement mismatches. Participants identify exaggerated assertions, unsupported arguments and misaligned supporting evidence within AI-generated content.
| Task | Identify overstated claims, unsupported arguments and misaligned evidence |
|---|---|
| Data | [TBD] |
| Shared task homepage | [link TBD] |
| Organizers | [TBD] |
| Registration | [link TBD] |
Shared Task Timeline
| Shared task call for participation | Tuesday | November 10, 2026 |
|---|---|---|
| Registration opens | Tuesday | November 10, 2026 |
| Training and development data release | Wednesday | November 25, 2026 |
| Test data release | Sunday | December 20, 2026 |
| System output submission deadline | Friday | December 25, 2026 |
| Evaluation results release | Wednesday | December 30, 2026 |
| System description paper deadline | Sunday | January 10, 2027 |
| Workshop day | [TBD] | [TBD], May 2027 |
Dates may be adjusted by the organizers of each track as needed.
All deadlines are 11:59 pm UTC-12h (anywhere on earth).
System Description Papers
Participating teams are invited to submit system description papers, which will be reviewed and, if accepted, presented in the shared task session at the workshop. Format and page limits: [TBD].