Method and device for fairness detection of artificial intelligence model

The invention provides a method and a device for fairness detection of an artificial intelligence model. The method comprises the following steps: outputting a plurality of fairness test items to be selected; determining at least one target item selected by the user from the plurality of to-be-selected fairness test items, and obtaining a target model input by the user and a sample set comprising a plurality of sample data; determining a plurality of sample attributes corresponding to the sample set; obtaining a sample attribute value of each piece of sample data in the sample set on each sample attribute, and obtaining a sample attribute value set corresponding to the sample set; generating a plurality of sample groups based on the sample attribute value set; the sample data in any sample group corresponds to a group of same sample attribute values; processing each piece of sample data in the plurality of sample groups by using the target model to obtain each prediction result corresponding to each piece of