生成式人工智能模型训练中的作品合理使用问题研究

打开文本图片集

[关键词］生成式人工智能模型训练合理使用文本与数据挖掘[中图分类号]G250.7[文献标志码］A［DOI]10.19764/j.cnki.tsgjs.20250616

[Abstract]neatitlitelgetoeluokstdelschlgspgtlsf coprightiritelotit modelsisnonexpresieandcanbeexplinedbyreasonableuse.Fomaneconomicperspetive，teuseofworksrecogidasfairusein thetrainingofgenatieartfiliteligeemodessotobofmarktfureduceialbasatenloaleical perspective，andpromotetecolgicalivationdeaepubliwefarefomapublicicperspetiVariouountriesepe diferentppasfogntitfitelgeelscssaalit is greaterdiscetiobsedsenaralysisthflexibleandrasoableusetesinJapanichdootrestritthpuosofusee notconducivetoprotectingtheinterestsofcopyrightowners;theEuropeanUnion’stextanddatamining（DM）exceptionrulesaremore conducivetoachevingabalanceofiterestsamongrelevantparties.WecanleamfromthepracticeofteEuropeanUnionandimproveTDM asanewsituationforthefairuseofcopyright，limitingtheaplicationofelementssuchasobject，purposeandbehavior.

[Keywords]Generativeartificial inteligence;Model training;Fairuse;Textand data mining

1问题的缘起

2022年11月，ChatGPT横空出世，引发全球热议。（剩余19014字）