Abstract: When you use your smartphone to browse news and videos, or go out for traveling, shopping, eating, etc., you are receiving product information recommendations from brands almost all the time in all aspects from screen time and hygiene habits to food and travel schedules. Therefore, artificial intelligence algorithmic recommendations, as a new generation of information technology applications that enable companies to outline user portraits more accurately, have become a commonly adopted method for companies to recommend relevant products or advertising information, permeating all aspects of life. However, will artificial intelligence algorithm recommendations increase consumers' brand favorability?
Based on the psychological resistance theory, this research starts from the perspective of consumers' psychological characteristics and uses scenario experiments and regression statistical analysis methods such as Bootstrap to study the impact of AI algorithmic recommendations on consumers' brand attitudes, and to explore in depth how the privacy and intrusive negative features of AI algorithmic recommendations affect consumers' attitudes toward brands, and whether consumers' brand attachment will improve their attitudes towards this influence. The results of the study are as follows: The privacy and intrusive features of AI algorithm recommendations have a significant negative impact on consumers' brand attitudes; The privacy and intrusiveness of AI algorithm recommendations lead to consumers' psychological resistance and thus affect consumers' brand attitude responses; Brand attachment has a significant positive moderating effect in the relationship between intrusive features and psychological resistance.
The theoretical contributions of this research are as follows: Firstly, it focuses on the negative effects of the current AI algorithm recommendations, and reveals reveals the mechanism of the impact of AI algorithm recommendations on consumers' psychological resistance. Secondly, it not only studies the moderating effect of the negative characteristics of brand attachment in AI algorithm recommendations on consumers' psychological resistance but also explores the boundary conditions of the influence of AI algorithm recommendations on consumers' brand attitudes. Thirdly, it conducts meaningful explorations of the mechanism of algorithm recommendations influencing consumers' brand response in the marketing field from the perspective of the relationship between AI algorithm recommendations and brands.
The research also has important management implications: First, enterprises should deeply cultivate the recommended content to avoid or reduce the impact of negative characteristics such as privacy and invasion when adopting AI algorithm recommendations for brand promotion. In the era of Web 3. 0, only by improving the connotation quality of recommended brand products and finding the content that can impress consumers the most can enterprises increase consumers' brand favorability and shape a good brand image. Second, it is necessary to improve the AI algorithm recommendation technology and form, give consumers more control over recommendations, set the selectivity of recommendations, and provide interpretable algorithm recommendation programs, to enhance the interactive experience and reduce psychological resistance. Moreover, enterprises should try to realize the change in AI algorithm recommendation strategy, and promote the change from “goods-looking-for-people” recommendation to “people-looking-for-goods” mode to attract consumers. Third, enterprises should attach great importance to cultivating consumers' brand attachment complex and segmenting consumer groups. For brand-attached consumer groups, especially “loyal” consumer groups, AI algorithm recommendations can be launched to promote quality content such as new features and performance of the brand products, to give consumers a new experience of the brand products, shorten the psychological distance with consumers, strengthen the emotional needs and shape the core competitiveness of the brand. However, it should not be used too much for objects without brand attachment, and blindly recommending them may backfire.

Key words: AI Algorithm Recommendation, Negative Characteristics, Psychological Resistance, Brand Attitudes, Brand Attachment

摘要： 利用人工智能算法推荐相关产品或广告信息已成为企业普遍采用的方式,然而人工智能算法推荐会增加消费者的品牌好感吗？本研究基于心理抗拒理论,从消费者的心理特征视角切入,利用情景实验以及Bootstrap等回归统计分析方法,研究人工智能算法推荐的隐私性和侵入性负面特征如何影响消费者对品牌的态度,以及消费者对品牌的依恋是否会改善其品牌态度。研究发现:人工智能算法推荐的隐私性和侵入性特征对消费者品牌态度产生显著负面影响;人工智能算法推荐的隐私性和侵入性特征会导致消费者的心理抗拒进而影响消费者的品牌态度;品牌依恋在侵入性特征与心理抗拒的关系中具有显著的正向调节作用。

关键词: 人工智能算法推荐, 负面特征, 心理抗拒, 品牌态度, 品牌依恋