Addressing Industry Adaptation Resistance in Combating Brand Deception: AI-Powered Technology vs. Revenue Sharing

Abstract

This paper studies a supply chain comprising a supplier, a third-party remanufacturer (TPR), and a retailer. The retailer sells both genuine and remanufactured products (i.e., Model O). Leveraging information advantages, the retailer may engage in brand deception by mislabeling remanufactured products as genuine to obtain extra profits (i.e., Model BD). AI-powered anti-counterfeiting technologies (AIT) (i.e., Model BA) and revenue-sharing contracts (i.e., Model C) are considered countermeasures. The findings reveal that (1) brand deception reduces (increases) sales of genuine (remanufactured) products, prompting the supplier (TPR) to lower (raise) wholesale prices. The asymmetric profit erosion effect highlights the gradual erosion of profits for the supplier, retailer, and TPR under brand deception. (2) The bi-interval adaptation effect indicates that AIT is particularly effective in industries with low adaptation resistance. When both the relabeling rate and industry adaptation resistance are low (high), Model BA (Model O) achieves a triple win. (3) Sequentially, when the industry adaptation resistance is low, AIT can significantly improve total profits, consumer surplus (CS), and social welfare (SW). Compared to Model BD, revenue-sharing offers slight advantages in CS but notable disadvantages in SW.

1. Introduction

In the remanufacturing market, retailers play multiple roles in the supply chain (SC), selling both genuine and remanufactured products [1]. For example, Miracle Automation Engineering, a retailer in Wuxi, sells new gearboxes from Aisin and Bosch, as well as remanufactured gearboxes from Weichai Remanufacturing Group (https://en.chinaconveyor.com/). In practice, both genuine and remanufactured products face demand uncertainty. Owing to abundant first-hand sales data, the retailer holds an information advantage over the supplier and the third-party remanufacturer (TPR) [2,3]. As a result, the retailer may engage in brand deception, relabeling remanufactured products as genuine to earn extra profits. However, brand deception in the remanufacturing industry poses significant safety risks, damages the reputation of genuine manufacturers, and erodes customer trust. Moreover, brand deception compromises SC transparency, triggers double marginalization, and leads to insufficient upstream supply competition and downstream product competition [3,4,5]. It negatively impacts SC’s profit performance by reducing demand for genuine products and increasing costs related to managing deception. Sequentially, consumer surplus (CS) is adversely affected as customers receive lower-quality products, undermining their perceived value and satisfaction. Social welfare (SW) is also compromised due to inefficiencies introduced by brand deception, resulting in a loss of trust in the market. Therefore, anti-counterfeiting strategies are urgently needed.

AI-powered anti-counterfeiting technology (AIT) utilizes advanced machine learning algorithms, image recognition, and blockchain technology (BCT) integration to guarantee product traceability and verify authenticity [6]. However, the introduction of new innovations and technologies in organizations often encounters initial resistance [7]. Moreover, the effectiveness of AITs varies across industries, as different sectors have unique demands, operational complexities, and levels of technological readiness, which influence their industry adaptation resistance.

Concretely, AI-powered image recognition technology detects counterfeit goods by analyzing product images, packaging, and unique design features. It can adapt to different angles and lighting conditions to ensure accuracy, particularly in sectors such as luxury goods [8]. In industrial practice, Lacoste has recently adopted AI-driven visual analysis, provided by Vrai AI, to verify product images and labels [9]. This ground-breaking AIT is ideal for companies across multiple industries, including jewelry [10], pharmaceutical [11], and electronics [12]. It is easy to implement and only requires a smartphone, making it accessible and cost-effective. As a result, the industry’s resistance to adaptation is low and the incentive to safeguard brand value is strong, making the integration of advanced image recognition technologies both feasible and effective. In contrast, BCT requires a more complex operational adaptation due to the need for a decentralized ledger and robust systems for data management. Blockchain’s integration into SCs demands a higher level of readiness across multiple stakeholders, as well as a significant upfront investment in infrastructure. This often results in higher industry adaptation resistance, particularly in auto parts remanufacturing with less technological readiness [13,14]. The operational capacity of industries like pharmaceuticals [15], where product traceability is a regulatory requirement, is better aligned with BCT, allowing for a smoother integration process. Thus, while AI image recognition technology can be rapidly deployed with minimal resistance in many industries, blockchain adoption tends to be more industry-specific and requires more robust operational capacity to be effective.

Revenue-sharing contracts are another widely accepted coordination mechanism that incentivizes SC members to align their behavior [16]. These contracts serve as a coordination tool, particularly when retailers engage in quantity-based competition and face demand dependent on costly retail efforts [17]. Studies have explored revenue-sharing contracts to promote information exchange and alleviate information leakage [18], as well as various coordination mechanisms in contract design [19,20,21,22,23], reverse SC [24,25,26], consignment contracts [27,28,29,30,31], and demand uncertainty [32,33], etc.

Our research question naturally arises: (1) How does brand deception impact SC members? (2) What are the strategic responses of the supplier and TPR under AIT and revenue sharing in combating brand deception? (3) Under what conditions do SC members prefer specific operational strategies in terms of total profit, CS, and SW?

To address the queries, this paper studies a SC comprising a supplier, a TPR, and a retailer. The retailer engages in brand deception by relabeling a portion of remanufactured products as genuine products to pursue extra profits (i.e., Model BD). One emerging solution is adopting AIT (i.e., Model BA), the other is a revenue-sharing contract to regulate retailer behavior (i.e., Model C). The main findings are as follows: (1) Brand deception negatively affects the supplier’s profits while both the TPR and retailer benefit at low deception levels, with the TPR being the biggest beneficiary. Brand deception results in decreased (increased) sales of genuine (remanufactured) products; thus, the supplier (TPR) tends to set a lower (higher) wholesale price. The asymmetric profit erosion effect was identified to highlight the dynamic process by which brand deception leads to the gradual erosion of profits for the supplier, retailer, and TPR. (2) For both the supplier and TPR, AIT adoption results in higher wholesale prices than in Model BD. In addition, order quantities surpass those in Model BD, provided that industry adaptation resistance remains below a certain threshold. The bi-interval adaptation effect indicates that AIT is particularly effective in industries with low adaptation resistance. When both the relabeling rate and industry adaptation resistance are low (high), Model BA (Model O) achieves a triple win. (3) Regarding social impact, brand deception can lead to a temporary increase in total profits and SW compared to Model O. However, once brand deception occurs, CS declines. Sequentially, when industry adaptation resistance is low, AIT can significantly improve the total profits, CS, and SW. Compared to Model BD, revenue sharing offers slight advantages in CS but significant disadvantages in SW.

This paper makes several significant contributions to the field of anti-counterfeiting strategies and brand management. First, it addresses a gap in the existing literature by focusing on proactive brand deception behavior initiated by retailers, rather than the more commonly studied passive counterfeit activities [34,35]. By analyzing the potential beneficiaries among TPRs, this research uncovers critical insights into the dynamics of brand deception and the power imbalance between SC members. Second, the paper assesses the effectiveness of AIT by examining its practical application in combating brand deception. The findings provide timely and valuable insights into the potential of AIT, highlighting the role of industry-specific adaptability and operational capacity in determining the success of AI solutions. Lastly, this research explores the broader societal implications of anti-counterfeiting strategies, examining how such initiatives affect total SC profits, CS, and SW. These findings offer actionable insights for engineering managers, underscoring the need for comprehensive strategies that balance economic interests with societal benefits. These contributions have practical applications in enhancing the effectiveness of anti-counterfeiting efforts and provide a foundation for future research in this critical area.

The remainder of this paper is organized as follows. Section 2 briefly reviews the related literature. Model settings are presented in Section 3. Comparative analysis is identified in Section 4. Section 5 extends the model to include the total profits, CS, and SW. Section 6 concludes the main findings, practical implications, and potential research directions. Notations, equilibrium outcomes, and proofs are summarized in the Supplementary Materials.

2. Literature Review

2.1. AI Technology Adoption and Adaptation

In recent years, the adaptation of AI technologies across various industries has garnered significant attention [36,37], as businesses and consumers increasingly recognize their potential to enhance efficiency, decision-making, and user experience.

Dwivedi et al. [38] explored the challenges, opportunities, and research agenda surrounding AI across multiple industries and domains. They highlighted the profound impact of AI technology on the future of industries and society, while also addressing the challenges and potential innovation opportunities. Sharma et al. [39] identified several key factors influencing retail customers’ adoption of AI-based autonomous decision-making systems, including effort expectancy, performance expectancy, facilitating conditions, and social influence, all of which positively affect AI adoption. Jiang et al. [40] explored how manufacturing firms can leverage AI to manage the associated opportunities and risks, driving iterative product innovation in the process. Through a three-case study of a typical Chinese manufacturing firm, they proposed key management principles for data management, experience integration, and intelligent device management. Olan et al. [41] studied the role of AI technologies and knowledge sharing in advancing consumer behavior. They found that when consumers share experiences in online communities, AI can promote changes in consumer attitudes and behaviors through learning and improvement, thereby driving product and service enhancements. Zhang et al. [42] examined the transformative impact of AI on human resource management practices using a quantitative descriptive approach. The study found that organizations using AI experienced significantly higher recruitment efficiency and employee productivity compared to those without AI integration. Bellis et al. [43] studied the barriers to consumer adoption of AI-driven autonomous shopping systems and proposed ways to overcome these barriers through the design of online and brick-and-mortar retail environments. Ameen et al. [44] studied how the integration of AI in shopping can enhance customer experience and proposed a theoretical model to analyze the roles of trust, commitment, perceived sacrifice, and service quality. Pillai et al. [45] investigated the factors that shape consumers’ intentions to shop at AI-powered automated retail stores. Hu et al. [46] explored how different types of artificial autonomy—sensing, thought, and action—affect user behavior in intelligent personal assistants. Chatterjee et al. [47] investigated the adoption of AI in higher education in India, applying specific models to understand stakeholders’ adoption behaviors. Fernandes et al. [48] examined the factors motivating consumers to adopt intelligent digital voice assistants in service encounters, with a particular focus on functional, social, and relational influences. Gursoy et al. [49] empirically tested an AI device use acceptance model in service encounters, exploring the motivations behind consumers’ acceptance of AI devices. Lin et al. [50] studied customers’ acceptance of using artificial intelligence robotic devices in hotel services, validating and extending the AI device use acceptance theory. Dong et al. [51] explored how users in the Chinese cultural context accept information technology innovations.

In the context of AI technology adoption, a common theme emerges regarding its acceptance across different stakeholders, such as customers, employees, and industries as a whole. Research consistently highlights that AI’s adoption faces varying degrees of resistance, with customers and employees often influenced by factors such as perceived usefulness, ease of use, and trust. Similarly, industries show differing levels of readiness based on their technological infrastructure, regulatory environment, and the urgency of addressing specific challenges. This paper identifies and examines this broader issue as general industry adaptation resistance. Notably, this paper represents the first to explore AITs within the framework of industry adaptation resistance, thereby creating a novel intersection in the literature between AI, anti-counterfeiting strategies, and operational management. By addressing this gap, the study contributes to the ongoing discourse in the OM field and offers new insights into the specific barriers and enablers influencing AI integration in the fight against counterfeiting across industries.

2.2. Anti-Counterfeiting Strategies

Anti-counterfeiting strategies have become a critical area of focus for businesses and regulators alike, as the proliferation of counterfeit goods continues to undermine brand integrity, consumer trust, and market competitiveness. Zhou et al. [52] investigated whether retail platforms and manufacturers should invest in anti-counterfeiting technologies. The study found that, under low production valuation, the payoff of anti-counterfeiting could be negative, and anti-counterfeiting measures might negatively affect CS and SW. Yao et al. [53] explored how traceability can address product label misconduct by comparing traceable and untraceable labeling systems. The study found that, while a traceable label system helps identify the responsible party, without proper management mechanisms in place, adopting a traceable system could result in higher costs and may even backfire. Yi et al. [54] investigated the impact of counterfeits on global SCs and how SC members can effectively take anti-counterfeiting actions. They analyzed the behavior of manufacturers and retailers and found that manufacturers prefer to incentivize retailers to combat counterfeits rather than take action themselves. The study showed that counterfeits can increase the SC’s profit even in the absence of network externalities, thereby alleviating double marginalization and benefiting the SC. Cho et al. [55] explored a scenario in which a legitimate retailer can control the proportion of counterfeit goods, analyzing how the brand owner responds to counterfeiting through adjustments in product quality and pricing strategies. Qian et al. [56] expanded upon the vertical differentiation model by incorporating elements of both searchable and experiential quality into their analysis. Zhang et al. [57] examined the role of implementing a direct channel as a strategy to mitigate the issue of counterfeits. Levi et al. [58] developed a comprehensive modeling framework to investigate economically motivated adulteration within agricultural SCs, analyzing farms’ strategic adulteration behaviors under different circumstances. Sun et al. [35] analyzed a retail platform’s optimal efforts in mitigating counterfeiting activities.

BCT has emerged as another popular strategy in combating counterfeiters [59,60,61,62]. For instance, BCT enhances traceability and authenticity in the food SC [63]. BCT improves tracking and combats counterfeit drugs in pharmaceuticals [64]. Platforms supported by BCT help verify the authenticity of diamonds and luxury items [8]. Centobelli et al. [65] designed a blockchain-based platform to improve the efficiency of recycling and remanufacturing processes. Pun et al. [62] adopted BCT for combating deceptive counterfeits while considering customer privacy. Wang et al. [34] examined whether retailers, aiming to protect their goodwill, should implement blockchain as an anti-counterfeiting strategy in scenarios involving asymmetric information on demand and product quality. Li et al. [66] analyzed the impact of blockchain technology adoption on an e-commerce SC, focusing on consumer privacy protection and sales mode selection. Holograms are also considered as an anti-counterfeiting strategy [67,68].

Our study differs from previous research in two key ways. First, we examine how retailers proactively initiate brand deception to gain excess profits, which leads to a different profit-diminishing trajectory compared to traditional competitive behaviors. This deviation highlights the unique motivations behind brand deception and its impact on the SC’s overall dynamics. Second, we approach the issue from the perspective of AI adaptation, exploring industry adaptation resistance and how varying operational adaptation capacities across industries influence the effectiveness of anti-counterfeiting strategies. Specifically, we analyze how these differences in adaptation capacities affect the adoption of AITs and revenue-sharing coordination, and their subsequent impacts on overall SC profitability, CS, and SW.

3. Model Setting

3.1. Original Model

Following Niu et al. [69], we developed an original model without brand deception, wherein the supplier and TPR (i.e., denoted by $s$ and $t$) offer genuine and remanufactured products with wholesale prices $w_s$ and $w_t$, respectively. Then, the retailer (i.e., denoted by $r$) places orders $q_s, q_t$ of the genuine and the remanufactured products to the supplier and TPR, respectively. Finally, the market retail price is $p_s, p_t$, respectively. The SC structures of the original model are illustrated in Figure 1. In addition, acronyms and notations are presented in

Table 1. Acronyms and notations.

Notations
${\pi }_j^i$	Firm $j$’s profit in model $i$, $i\in \left\{O,BD,BA,C\right\}$ and $j\in \left\{s,t,r\right\}$
$p_j^i$	Firm $j$’s retail price in model $i$, $i\in \left\{O,BD,BA,C\right\}$ and $j\in \left\{s,t,r\right\}$
$a$	The demand potential
$\theta$	The demand potential after AIT
$\epsilon$	The demand uncertainty level
$\alpha$	The relabeling rate of counterfeit products under brand deception
$\delta$	Brand’s reputation retention rate with deception
$\lambda$	The revenue-sharing rate
$\tau$	Threshold of relabeling rate
$c$	The industry adaptation resistance
${CS}^i$	Consumer surplus in model $i$, $i\in \left\{O,BD,BA,C\right\}$
${SW}^i$	Social welfare in model $i$, $i\in \left\{O,BD,BA,C\right\}$
$w_j^i$	Firm $j$’s wholesale price in model $i$, $i\in \left\{O,BD,BA,C\right\}$ and $j\in \left\{s,t,r\right\}$
$q_j^i$	Firm $j$’s sale quantity in model $i$, $i\in \left\{O,BD,BA,C\right\}$ and $j\in \left\{s,t,r\right\}$

and the game is solved by backward induction and the sequence is illustrated in Figure 2.

Following [3,4,70], the inverse demand functions of the supplier and the retailer are: $p_s^O={\epsilon }_{1}a-q_t^O-q_s^O; p_t^O={\epsilon }_{2}a-q_t^O-\delta q_s^O$. The corresponding profit function of the supplier, TPR, and the retailer are: ${\pi }_s^O=w_s^O\delta q_s^O; {\pi }_t^O=w_t^O{q}_t^O; {\pi }_r^O=\left(p_s^O-w_s^O\right)\delta q_s^O+\left(p_t^O-w_t^O\right)q_t^O$.

3.2. Brand Deception

We consider a situation in which the genuine products and the remanufactured products encounter challenges associated with demand uncertainty. Meanwhile, the retailer holds an information advantage over the supplier and the TPR owing to abundant first-hand sales data. For retailers, there is a profit margin in taking advantage of this information gap. Therefore, the retailer may resort to a risky, unfair competition strategy—relabeling a portion (i.e., relabeling rate $\alpha$) of remanufactured products as genuine to pursue heightened profits, constituting brand deception. Brand deception is bound to cause reputational damage, which results in quantity discounts $(1-\delta )$ for the supplier and the retailer faces a profit reduction of $(1-\delta )$ as a penalty. Here, $\delta$ represent the reputation retention rate. Consequently, the reputational damage will certainly not be too low or even reach 0 and the relabeling rate cannot be too low, excessive, or even reach 100%; such a condition is not practical or without implications. In this regard, we assume there is a threshold $\tau \in ({\displaystyle \frac{1}{2}},1)$ which restricts $\alpha \in \left[0,\tau \right]; \delta \in (\tau ,1]$. To explore general situations, we normalize $\tau$ to 0.7 following empirical thresholds used in the related literature [61,71], which aligns with practical enforcement cutoffs observed in platform monitoring systems.

Moreover, ${\epsilon }_{i}a$ denotes the uncertain demand and the deterministic part is $a$, which is normalized to 1, and the random part ${\epsilon }_i\in \left\{\mathrm{0,1}\right\}$ follows a normal distribution with mean $\mu =1$. We assume that the TPR faces a smaller demand variance $\eta {\sigma }^2$ than the supplier ${\sigma }^2$, where $\eta <1; {\sigma }^2<1$. Note that the supplier and TPR’s responses are based on the expected demand and the retailer’s response is based on full distributional information. Therefore, only the retailer’s expected profit expression includes distributional parameters (i.e., $\eta ,{\sigma }^2$).

Finally, the inverse demand functions of the supplier and the retailer are: $p_s^{BD}={\epsilon }_{1}a-q_s^{BD}-q_t^{BD}; p_t^{BD}={\epsilon }_{2}a-q_t^{BD}-\delta q_s^{BD}$. Correspondingly, the profit functions of the supplier and TPR are: ${\pi }_s^{BD}=w_s^{BD}\delta q_s^{BD}; {\pi }_t^{BD}=w_t^{BD}{q}_t^{BD}$. The retailer’s profit is ${\pi }_r^{BD}=\left(p_s^{BD}-\left(1-\delta \right)\right)(q_s^{BD}+\alpha q_t^{BD})+p_t^{BD}(1-\alpha )q_t^{BD}-w_s^{BD}{q}_s^{BD}-w_t^{BD}{q}_t^{BD}$, including the counterfeit product’s profit $\alpha q_t^{BD}$.

3.3. Brand Deception Under AIT

Intuitively, brand deception cannot persist for long due to the inevitable damage to reputation and diminishing profit margins for retailers. When the costs of deception outweigh the benefits, retailers are likely to abandon deceptive practices. To counteract brand deception, both suppliers and TPRs may adopt AITs, which encompass advanced machine learning algorithms, image recognition, and blockchain integration. However, even though these technologies enable real-time detection of counterfeit products and ensure traceability for both genuine and remanufactured products, the resistance of different industries to adopting and applying emerging technologies varies. Thus, we introduce the industry adaptation resistance $c$ which varies across sectors, as different industries possess unique operational complexities, technological readiness, and regulatory environments. The effectiveness of AITs depends on how well each industry can integrate these solutions. Industries with lower adaptation resistance can more readily implement and benefit from these technologies, while those with higher adaptation resistance may face challenges in achieving similar outcomes.

When AIT is adopted, the deterministic demand expands to $\theta$, though as indicated in [15], this type of expansion cannot exceed $\theta <{\displaystyle \frac{3}{2}}$. For both suppliers and TPRs, $c$ reflects the industry’s technological investment and the specific challenges it faces in implementing these solutions. With these adjustments, the inverse demand functions are as follows: $p_s^{BA}={\epsilon }_1\theta -q_s^{BA}-q_r^{BA}; p_t^{BA}={\epsilon }_2\theta -q_t^{BA}-\delta q_s^{BA}$. Consequently, the profit functions of the SC members are: ${\pi }_s^{BA}=\left(w_s^{BA}-c\right)\delta q_s^{BA}; {\pi }_t^{BA}=\left(w_t^{BA}-c\right)q_t^{BA}; {\pi }_r^{BA}=\left(p_s^{BA}-\left(1-\delta \right)\right)\left(q_s^{BA}+\alpha q_t^{BA}\right)+p_t^{BA}\left(1-\alpha \right)q_t^{BA}-(w_s^{BA}+c)q_s^{BA}-(w_t^{BA}+c)q_t^{BA}$.

3.4. Revenue-Sharing Coordination

Model C differs from the two models mentioned in that the supplier offers a revenue-sharing contract. Both the supplier and the TPR have the same inverse demand functions as in Model BD: $p_s^C={\epsilon }_{1}a-q_s^C-q_t^C; p_t^C={\epsilon }_{2}a-q_t^C-\delta q_s^C$. Since revenue-sharing contracts are a widely adopted and traditional method, we assume that the industry adaptation resistance to this approach is $0$ in this paper. Notably, there are many types of revenue-sharing contracts [17,72,73]. This paper assumes the retailer shares the sold genuine product’s revenue at the rate $\lambda$ [17]. Consequently, the supplier’s profits include the basic profit $w_s^C\delta q_s^C$, the shared normal part, and the shared brand deception part $\lambda p_s^C\left(q_s^C+\alpha q_t^C\right)$.

The retailer’s profits include the profit from genuine products $(1-\lambda )p_s^C{q}_s^C$, the brand deception part $(1-\lambda )p_s^C\alpha q_t^C$, and the profit punishment part $\left(1-\delta \right)(q_s^C+\alpha q_t^C)$. Note that the profit of the remanufactured product is decreased to $p_t^C(1-\alpha )q_t^C$. The profit functions of SC members are: ${\pi }_s^C=\lambda p_s^C\left(q_s^C+\alpha q_t^C\right)+w_s^C\delta q_s^C; {\pi }_t^C=w_t^C{q}_t^C; {\pi }_r^C=\left(\left(1-\lambda \right)p_s^C-\left(1-\delta \right)\right)(q_s^C+\alpha q_t^C)+p_t^C(1-\alpha )q_t^C-w_s^C{q}_s^C-w_t^C{q}_t^C$. To ensure the existence of the equilibrium and positive results in Model C, we assume $0<\lambda <2\sqrt{{\displaystyle \frac{(\delta -1)(\alpha \delta -1)}{{(\alpha -1)}^2}}}+{\displaystyle \frac{(1+\alpha )(1-\delta )}{\alpha -1}}$ and $\lambda <{\lambda }_1$.

4. Model Analysis

4.1. Analysis of Wholesale Price

By comparing the wholesale price decisions between Model O and Model BD, Proposition 1 is obtained. Then, we compare the wholesale price decisions between Model BD and Model BA, and Proposition 2 is obtained.

Proposition 1.

The size relationship of wholesale price between Model O and Model BD:

(1)

For supplier, $E\left[w_s^O\right]>E\left[w_s^{BD}\right]$;

(2)

For TPR, $E\left[w_t^O\right]<E\left[w_t^{BD}\right], if \left\{\begin{array}{c}0<\alpha <0.126; \tau <\delta <1\hfill \\ 0.126<\alpha <0.2; {\delta }_1<\delta <1\hfill \end{array}\right.$.

The presence of brand deception is always detrimental to the supplier. In Model O, the supplier is unknowingly involved in invisible competition caused by the retailer’s brand deception. Retailer brand deception behavior involves rebranding counterfeit products to occupy part of the market share of the supplier, resulting in a loss of marginal profits. To address this challenge, the supplier must employ a strategic pricing approach that accounts for both channels’ profits, considering the potential impact of brand deception by the retailer. Additionally, implementing robust monitoring mechanisms to detect and mitigate instances of counterfeit products can safeguard the supplier’s market share and preserve overall profitability.

Confronting that, the supplier has three strategies: (1) increasing $w_s^{BD}$ to enhance profit margins; (2) decreasing $w_s^{BD}$ to boost order quantities from the retailer; or (3) promoting the BCT adoption or offering a revenue-sharing coordination contract. The first strategy could be detrimental to the supplier because the increased wholesale price may lead to rampant brand deception by the retailer since the quantity competition further decreases. Differently, the second choice could lead to fierce quantity competition with remanufactured goods in the retailer channel; thus, lowering prices is a conservative and effective measure that suppliers can take. The third choice aims to construct a transparency channel and break information asymmetry with robust monitoring mechanisms.

Proposition 2.

For both supplier and TPR, the wholesale price in Model BA is higher than in Model BD, $0<E\left[w_j^{BD}\right]<E\left[w_j^{BA}\right],j\in \left\{s,t\right\}$.

Proposition 2 highlights that when AI-powered anti-counterfeiting technology is adopted, both the supplier and TPR set higher wholesale prices compared to Model BD. This is because AITs, regardless of the specific implementation across industries, enhance product authenticity and transparency, thereby increasing perceived product value. The resulting confidence enables suppliers and TPRs to set higher prices. While industry adaptation resistance varies, the fundamental effect of increased wholesale prices persists universally, driven by the intrinsic value AI adds in combating brand deception and improving SC reliability.

4.2. Analysis of Order Quantities

Proposition 3.

The size relationship of order quantities between Model O and Model BD:

(1)

For supplier, $E\left[q_s^O\right]>E\left[q_s^{BD}\right]$;

(2)

For TPR, $E\left[q_t^O\right]<E\left[q_t^{BD}\right]$.

Despite in-store competition at retailers, the profit margins stemming from brand deception have the potential to augment overall sales for the retailer. Concurrently, this deceptive strategy may result in an inevitable reduction in the sale quantity of new products through the supplier channel, particularly if market demand remains stable. As remanufactured products vie with genuine products within retailers’ stores, the heightened sales quantity of remanufactured products directly contributes to an escalation in wholesale prices $w_r^O$. The intricate interaction between remanufactured and new products, intensified by the in-store competition at the retailer, augments the demand for TPRs. Consequently, this surge exerts upward pressure on wholesale prices, ultimately fortifying the profitability of the TPR. Consequently, the advantages accrued by the TPR underscore its gains stemming from the retailer’s brand deception.

Proposition 4.

For both supplier and TPR, there exists a threshold of BCT adoption cost, such that $0<E\left[q_j^{BD}\right]<E\left[q_j^{BA}\right],j\in \left\{s,t\right\},if c<{\displaystyle \frac{\theta -1}{2}}$.

There is a critical cost threshold (i.e., $c<{\displaystyle \frac{\theta -1}{2}}$) for adopting AITs beyond which order quantities for both genuine and remanufactured products are higher than in Model BD. This indicates that beyond simply reducing deception, effective AITs can stimulate demand by enhancing consumer confidence in product authenticity, thus driving larger order quantities, and benefiting the entire SC.

4.3. Analysis of Profit

By comparing the profit difference between Model O and Model BD, Proposition 5 is obtained. Then, we obtain Proposition 6 to investigate the effectiveness of AITs.

Proposition 5.

The size relationship of profit between Model O and Model BD:

(1)

For supplier: $E\left[{\pi }_s^O\right]>E\left[{\pi }_s^{BD}\right];$

(2)

For TPR: $E\left[{\pi }_t^O\right]<E\left[{\pi }_t^{BD}\right], if \left\{\begin{array}{c}0<\alpha <0.274; \tau <\delta <1\hfill \\ 0.274<\alpha <0.36; {\delta }_3<\delta <1\hfill \end{array}\right.;$

(3)

For retailer: $E\left[{\pi }_r^O\right]<E\left[{\pi }_r^{BD}\right], if \left\{\begin{array}{c}0<\alpha <0.208; \left\{\begin{array}{c}\tau <\delta <{\delta }_4\hfill \\ {\delta }_4<\delta <{\delta }_5\hfill \end{array}\right.\hfill \\ 0.208\le \alpha \le 0.229; \tau <\delta <{\delta }_5\hfill \\ 0.229<\alpha <\tau ; \tau <\delta <{\delta }_5\hfill \end{array}\right.$.

First, the higher relabeling rate led to more damage to their profits for all SC members. Second, once a retailer chooses to pursue a brand deception strategy, the supplier suffers from reputational damage and order quantity, thus maintaining sustained profit reduction. Third, as per Proposition 3, the sale quantity of the remanufactured product also increases. Interestingly, in the process of increasing the relabeling rate, the relative increase interval in profits, of which TPRs are the biggest beneficiary, will be greater than that of retailers. However, this benefit is short-sighted and will do long-term damage to the brand image of suppliers’ genuine products. In addition, the TPR’s remanufactured products will also suffer reputational damage due to nepotism. Therefore, the potential profit reduction encourages suppliers to devote themselves to achieving SC transparency and coordination to ensure stable market sales and avoid unfair invisible competition.

Furthermore, as the relabeling rate increases in Figure 3, the supplier, retailer, and TPR experience an Asymmetric Profit Erosion Effect. The supplier’s profit is negatively impacted from the outset and worsens progressively with higher counterfeiting rates. The retailer initially benefits from low counterfeiting rates, but its profit turns negative after a certain threshold and continues to decline with further increases in the counterfeiting rate. In contrast, the TPR demonstrates greater resilience, remaining profitable under higher counterfeiting rates; however, it eventually suffers losses as the counterfeiting rate rises further. This effect illustrates a dynamic progression in the brand deception game: from an initial phase where the supplier incurs losses while the retailer and TPR gain profits (Lose–Win–Win), to an intermediate phase where the supplier and retailer both face losses while the TPR remains profitable (Lose–Lose–Win), and finally to a stage where all three parties suffer losses (Lose–Lose–Lose). The asymmetry arises from differences in the roles and profit structures of the three parties, leading to distinct thresholds of vulnerability and varying degrees of profit erosion.

Proposition 6.

The size relationship of profit between Model BD and Model BA:

(1)

For supplier, $E\left[{\pi }_s^{BA}\right]>E\left[{\pi }_s^{BD}\right], if \left(c<{\displaystyle \frac{\theta -1}{2}}|| c>{\phi }_7\right)$

(2)

For TPR, $E\left[{\pi }_t^{BA}\right]>E\left[{\pi }_t^{BD}\right], if \left(c<{\displaystyle \frac{\theta -1}{2}}|| c>{\phi }_8\right)$

Similar to the inefficient interval phenomena identified in prior studies [74], we uncover a bi-interval adaptation effect in Proposition 6 and Figure 4. How the industry adaptation resistance shapes the economic feasibility of AIT adoption is reflected. For suppliers, adopting AI is profitable when the adoption cost ccc falls below ${\displaystyle \frac{\theta -1}{2}}$ or exceeds a benefit–cost threshold ${\phi }_5$. The intermediate range $({\displaystyle \frac{\theta -1}{2}},{\phi }_5)$ represents an inefficient interval where industry adaptation resistance arises, as the potential gains from mitigating brand deception fail to justify the investment, leading to hesitancy in adoption.

For TPRs, the cost effect of AIT adoption mirrors that of suppliers, but with notable differences due to the bi-interval adaptation effect and industry adaptation resistance. TPRs realize benefits only when costs exceed a higher threshold ${\phi }_6$, reflecting their resilience to retailer deception and potential advantages from brand deception. The inefficient interval, spanning $({\displaystyle \frac{\theta -1}{2}},{\phi }_6)$, is broader, indicating that TPRs are less sensitive to costs but face stronger resistance due to limited net profitability gains. This highlights their need for either lower adoption costs or greater benefits to justify investment.

For industries with high operational adaptability, such as luxury goods and fashion, the inefficiency interval for AI adoption costs is relatively narrow. These sectors, characterized by rapid market changes and digital-first consumer engagement, can integrate AITs seamlessly to enhance transparency and counteract brand deception. Their flexible operations and emphasis on innovation allow them to achieve a favorable cost–benefit balance for AI implementation. In contrast, industries with low operational adaptability, like traditional manufacturing and retail, encounter broader inefficiency intervals. These sectors often struggle with legacy systems, rigid operational frameworks, and slower adoption of digital tools, which elevate the costs of AI adoption. As a result, the economic viability of AI in these industries requires either a significant reduction in adoption costs or substantial external support, such as government incentives or technological partnerships, to overcome these barriers. This disparity highlights the critical role of industry-specific strategies in ensuring that the benefits of AIT are realized across diverse sectors.

We conducted an extensive numerical study due to the complexity of deriving analytical comparisons. The retailer’s response was evaluated using complete distributional data. Our analysis involved six variables: $\alpha ,\delta ,\theta ,\eta ,\sigma ,c$, with values specified in

Table 2. Parameter values in Observation 1.

Parameters	Lower Bound	Upper Bound	Step	Count
$\alpha$	0.1	0.7	0.1	7
$\delta$	0.7	1	0.025	13
$c$	0.1	1	0.15	7
$\theta$	1.1	1.5	0.025	17
$\eta$	0.5	0.9	0.025	17
$\sigma$	0.1	0.9	0.025	33

. The study resulted in $\mathrm{6,075,069}$ unique combinations of these variables $7\times 13\times 7\times 17\times 17\times 33$. Computations were performed in Python 3.8 and R 4.3.3, and the results were aggregated by the $(\alpha ,c)$ pair. Figure 5 visualizes these findings, presenting the ratio of combinations favoring Model BA over Model BD under constant $(\alpha ,c)$ values, and Observation 1 is obtained.

Observation 1.

As the adaptation resistance $c$ increases, the retailer’s incentive to choose AIT initially decreases and subsequently increases in $\alpha$.

Observation 1 shows that (1) retailers may not always reject the introduction of AIT. Their decision to adopt AI or continue brand deception is influenced by both the adaptation resistance $c$ and the relabeling level $\alpha$. When resistance is low, AI adoption is more attractive as it enhances traceability and consumer trust without excessive resistance, thus reducing the incentives for deception. (2) As resistance increases, the industry’s adaptation resistance, particularly from suppliers and TPRs, can make the initial investment in AI less justifiable, especially when deception levels are high. (3) At high resistance levels, the long-term benefits of AI, such as reduced deception risk and improved reputation, may make AI adoption more attractive. Retailers must carefully manage adaptation resistance and develop strategic plans, particularly when deception is high, to ensure a profitable and sustainable transition to AIT.

4.4. Profit Preference Analysis

To examine the profit preference of participants, we set the same parameter value with $(\lambda =0.01; \theta =1.5; \eta =0.7; \sigma =0.8)$. Figure 6a indicates whether industries with high operational adaptability, such as luxury goods and fashion, and Model BA’s bi-interval adaptation effect can effectively leverage AIT. Specifically, when the relabeling rate is very low and the reputational damage is minimal, these industries can easily use AIT such as image recognition and QR codes to enhance product traceability and build consumer trust.

Figure 6b confirms the bi-interval adaptation effect of Proposition 6. When the adaptation resistance satisfies ${\displaystyle \frac{\theta -1}{2}}<c<{\phi }_7$ and the sharing rate is relatively low, although supplier might extract greater profits from AIT and revenue sharing when reputation damage is low, the supplier is more inclined to eliminate brand deception. Figure 6c shows that although traditional manufacturing industries typically face high industry adaptation resistance and low operational adaptability, the integration of AIT can significantly expand the profit margins for suppliers. This is because these technologies enhance SC transparency and efficiency, reduce the risk of brand deception, and strengthen consumer trust in product quality and brand reputation. Such advancements offset the high initial costs, enabling suppliers to achieve substantial economic gains in the long run. For Model C, the supplier’s profit increase is brought about by sharing the excess profits from brand deception.

Figure 7 confirms that the TPR has no motivation to choose Model C rather than Model BD. Figure 7a indicates that when the adaptation resistance is very low, the effectiveness of AIT is extremely high, enabling the TPR to take advantage of the transparency and traceability provided by AIT, thereby reducing brand deception and increasing market demand and profits. At this time, the gain effect of AIT exceeds its adaptation resistance, resulting in an overall increase in the profits of the TPR. However, when the degree of brand deception is high, the TPR will not be able to obtain high profits from brand deception, which is consistent with Proposition 5.

Figure 7b confirms Proposition 6. When the adaptation resistance satisfies ${\displaystyle \frac{\theta -1}{2}}<c<{\phi }_8$, the TPR will completely give up AIT. At this time, when the deception level is very low, the TPR may be more willing to maintain the existing fraudulent operation and is more inclined to adopt Model BD to be the biggest beneficiary of brand deception. As the deception level increases, the reputation of the TPR will also be damaged. It can be said that the overall reputation of the SC has been destroyed, and the TPR will return to Model O. Figure 7c shows that when the adaptation resistance is very high, AI-powered technology is not enough to eliminate the TPR’s tendency toward brand deception. For the TPR, AIT is particularly suitable for situations where reputational damage is high and the deception level is low.

Figure 8 shows the effectiveness of AI-powered technology in dealing with brand deception. Figure 8a shows that when the adaptation resistance is very low, suppliers can easily integrate AIT to improve product tracking and verification capabilities. This allows consumers to confirm that the product they are purchasing is authentic. At the same time, low-resistance AIT solutions can significantly reduce the risk of counterfeit products, thereby protecting the interests of brands and consumers and reducing potential economic losses. Figure 8b illustrates that as the adaptation resistance increases, suppliers and TPRs may face the inefficiency range of AIT. Moreover, as the deception level increases, retailers will not benefit from brand deception and therefore tend to choose Model O. Figure 8c points out that when adaptation resistance is very high, it is still possible for retailers to obtain excess profits by continuing brand deception. At the same time, they may also be inclined to obtain higher profits through revenue sharing. However, the effect of revenue sharing is only apparent when reputational damage is high and deception is low.

Overall, Figure 9 reveals the trade-off between industry adaptation resistance and technological benefits. When the adaptation resistance is low and the relabeling rate is low, the traceability and trust advantages of AIT achieve a triple win, whereas at higher adaptation resistance and relabeling rates, eliminating brand deception becomes more attractive. In Figure 9a, with low adaptation resistance $c=0.02$, Model BA achieves a triple win due to its cost-effectiveness. At low adaptation resistance, the implementation cost of AIT is significantly reduced, and the gains from enhanced traceability and consumer trust far outweigh the investment. The low cost and high effectiveness of AI quickly weaken the appeal of brand deception, creating a mutually beneficial scenario for suppliers, third-party remanufacturers, and retailers.

Figure 9b corresponds to a higher industry adaptation resistance range $c\in \left[\mathrm{0.6,0.9}\right]$. In this context, Model O achieves a triple win, concentrated in regions with moderate relabeling rates and moderate-to-high reputation retention. The core reason is that, as adaptation resistance increases, the initial cost and barriers to implementing AIT rise, making it harder for retailers to gain short-term benefits. Under such conditions, eliminating brand deception becomes more economically rational, as it avoids the high costs associated with technological adaptation while fostering trust and collaboration to achieve a triple win.

5. Extensions

Brand deception poses a serious threat in the retail sector, undermining consumer trust and potentially distorting market competition, thus harming the principles of fair trade. As BCT increasingly becomes integrated to ensure product authenticity and traceability, examining its impact on reducing instances of brand deception is of paramount importance. Additionally, evaluating the role of revenue sharing in preventing brand deception is crucial for a comprehensive understanding of market mechanisms and enhancing overall SW. Hence, this paper focuses on how these mechanisms can improve CS and SW, propelling a more equitable and sustainable market environment. Intuitively, the total profit of the SC is $T^i={\pi }_s^i+{\pi }_t^i+{\pi }_r^i; i\in (O,BD,BA,C)$. Note that $c\left(q_s^{BA}+q_t^{BA}\right)$ denotes the total cost of AI-powered technology implementation and $\alpha \left(1-\delta \right)\left(q_s^C+q_t^C\right)$ denotes the losses caused by brand deception. Then, following [75], the CS and SW functions in the four models are defined in

Table 3. Definition of CS and SW.

	${\mathit{C}\mathit{S}}^{\mathit{i}},\mathit{i}\in \left\{\mathit{O},\mathit{B}\mathit{D},\mathit{B}\mathit{A},\mathit{C}\right\}$	${\mathit{S}\mathit{W}}^{\mathit{i}},\mathit{i}\in \left\{\mathit{O},\mathit{B}\mathit{D},\mathit{B}\mathit{A},\mathit{C}\right\}$
Model O	${\displaystyle \frac{\left({\epsilon }_{1}a-p_s^O\right)q_s^O+\left({\epsilon }_{2}a-p_t^O\right)q_t^O}{2}}$	${CS}^O+{\pi }_s^O+{\pi }_t^O+{\pi }_r^O$
Model BD	${\displaystyle \frac{\left({\epsilon }_{1}a-p_s^{BD}\right)\left(q_s^{BD}+\alpha q_t^{BD}\right)+\left({\epsilon }_{2}a-p_t^{BD}\right)\left((1-\alpha )q_t^{BD}\right)}{2}}$	${CS}^{BD}+{\pi }_s^{BD}+{\pi }_t^{BD}+{\pi }_r^{BD}$
Model BA	${\displaystyle \frac{\left({\epsilon }_1\theta -p_s^{BA}\right)\left(q_s^{BA}+\alpha q_t^{BA}\right)+\left({\epsilon }_2\theta -p_t^{BA}\right)\left((1-\alpha )q_t^{BA}\right)}{2}}$	${CS}^{BA}+{\pi }_s^{BA}+{\pi }_t^{BA}+{\pi }_r^{BA}-c\left(q_s^{BA}+q_t^{BA}\right)$
Model C	${\displaystyle \frac{\left({\epsilon }_{1}a-\left(1-\lambda \right)p_s^C\right)\left(q_s^C+\alpha q_t^C\right)+\left({\epsilon }_{2}a-p_t^C\right)\left((1-\alpha )q_t^C\right)}{2}}$	${CS}^C+{\pi }_s^C+{\pi }_t^C+{\pi }_r^C-\alpha \left(1-\delta \right)\left(q_s^C+q_t^C\right)$

:

5.1. Total Profit Analysis

After defining the CS and SW, we focus on the total profit under the four models. Figure 10 shows that (1) the total profit in the three models other than the original Model O shows a decreasing trend with an increase in industry adaptation resistance $c$. First, AIT incurs significant implementation expenses, and when costs rise, the financial benefits for the total SC diminish. Beyond a certain cost threshold, the financial advantages of BCT may no longer outweigh the expenses, resulting in total profit being lower than Model O (i.e., $T^{BA}<T^O$). Higher relabeling rates indicate a greater prevalence of brand deception, which severely damages the SC’s reputation. AIT helps distinguish genuine remanufactured products from counterfeits, thus preventing deceptive practices. This increases customer trust in remanufactured products, benefiting the TPR by increasing demand and profitability. The value of AIT is most significant when $\alpha$ is high and $c$ is low. (2) In Figure 10b, the total profit in all four models shows a decreasing trend when the degree of uncertainty in the SC increases. Intuitively, companies face higher costs and operational challenges, leading to a decrease in total profit across different models. This trend underscores the importance of effective risk management and SC optimization strategies to mitigate the adverse effects of uncertainty.

Figure 10 indicates that the total profit in Model C is lower than in Model BD. Higher $\lambda$ reduces the retailer’s incentive to focus on product quality and increases the likelihood of deception. Thus, the overall SC profitability in Model C is reduced. Higher $\alpha$ negatively affects the retailer’s reputation and reduces consumer trust. In Model C, the revenue-sharing mechanism is less effective in preventing deceptive practices at higher $\alpha$, reducing total profit.

When the relabeling rate is low, the SC is able to obtain greater total profits. When brand deception occurs, both the TPR and the retailer will gain additional profits, and the sum of the profits is greater than the supplier’s profit loss. When the relabeling rate increases, the gain effect of this brand deception will no longer exist because the reputational damage at this time has accumulated to the point that the SC loses profitability. This also confirms the finding in Proposition 4 regarding TPRs and retailers’ profit advantages.

AIT not only enhances transparency but also acts as a deterrent against brand deception, leading to increased trust and reduced incidents of counterfeit products. AI-powered technology ensures that product origins and quality are verifiable, which in turn attracts consumers willing to pay a premium for guaranteed authenticity. This shift results in a higher demand for genuine products, which directly contributes to increased profitability throughout the SC. Additionally, the implementation of AIT might initially increase operational costs, but these are offset by the long-term gains in consumer trust and reduction in losses due to counterfeit returns.

5.2. Consumer Surplus Analysis

Figure 11 shows that (1) when brand deception occurs, $CS$ shows a decreasing trend with an increase in adoption cost $c$. Higher $c$ directly reduces the net CS since these costs are often passed down to consumers through increased product prices. Additionally, high $c$ may deter SC members from implementing AIT, reducing its impact on improving transparency and quality. (2) Intuitively, $CS$ shows an increasing trend with an increase in brand deception damage factor $\delta$. It is not difficult to understand that when reputational damage is low, the negative effects caused by brand deception are naturally smaller. Higher brand deception damage causes significant reputational losses, reducing consumer trust in genuine and remanufactured products. (3) ${CS}^{BA}-{CS}^{BD}$ remains stable with an increase in relabeling rate $\alpha$. Higher $\alpha$ indicates more frequent brand deception practices, affecting both genuine and remanufactured products. However, with AIT, transparency and traceability are maintained regardless of relabeling practices, providing consumers with accurate product information. As a result, CS remains relatively stable across varying $\alpha$. AIT helps mitigate brand deception and restore consumer trust through improved transparency and traceability.

In addition, intuitively, when brand deception occurs, $CS$ shows a decreasing trend with an increase in relabeling rate $\alpha$. Higher $\alpha$ values indicate more brand deception. Implementing revenue sharing effectively reduces this behavior, increasing consumer trust and market demand, thereby resulting in a modest increase in CS. Second, the introduction of revenue sharing will not significantly improve CS. As the revenue-sharing rate $\lambda$ increases, the incentive alignment among SC members becomes more evident, leading to collaboration in reducing brand deception and improving product quality, which results in a slight increase in CS. In addition, at higher levels of $\lambda$, although the supplier’s profit share increases, the overall market demand and consumer trust improvement can still result in a modest increase in CS. Third, lower $\delta$ values indicate greater reputational damage, so reducing brand deception through revenue sharing, while having a limited impact, still positively affects CS.

Figure 11 showcases how AIT enhances CS by ensuring the authenticity of products, thereby increasing consumer confidence and willingness to engage in the market. As AIT provides irrefutable proof of product quality and origin, it reduces the uncertainty consumers face when making purchasing decisions. This reduction in uncertainty typically results in a willingness to pay more for assured quality, which directly boosts CS. The revenue-sharing coordination model aligns the economic incentives of all SC members, which can slightly increase CS. This model encourages all involved parties to focus on delivering higher-quality products and maintaining integrity in product descriptions, thus increasing the overall value received by consumers. Particularly in cases of low reputational damage and high relabeling rate, revenue-sharing coordination shows a modest advantage in enhancing CS.

5.3. Social Welfare Analysis

Note that market expansion (i.e., $\theta =1.5$) is introduced by AIT and only affects the performance of Model BA. Thus, a decline in market demand will bring about a significant decline in SW. In addition, the switch from solid to dashed blue lines represents adoption cost $c$ increases, highlighting the trade-off between the benefits of AIT in enhancing the market size and its costs. Intuitively, the $SW$ in three models show a decreasing trend with an increase in $\alpha$. Higher $\alpha$ implies more frequent brand deception, which erodes consumer trust and reduces the SW in Figure 12. ${SW}^{BA}$ reaches its highest level when the industry adaptation resistance $c$ is low, market expansion potential $\theta$ is high, and uncertainty level is low. With the transparency and traceability provided by AIT, consumers have increased trust in products. AIT can significantly increase market demand, increase the profits of SC members, and increase CS accordingly, thereby improving the overall SW. The worst situation is the combination of high $c$, low $\theta$, and high uncertainty level, when the demand expansion brought by AIT is not enough to cover its costs, and SW is low or even negative.

${SW}^C$ is lower than ${SW}^{BD}$ because revenue-sharing coordination involves additional administrative and operational costs to manage and monitor the contracts. When the relabeling rate $\alpha$ is low, these costs might be justified by the reduction in brand deception. However, as $\alpha$ increases, the marginal benefit of reducing brand deception diminishes, while the costs remain high, leading to lower overall SW. Regarding ${SW}^{BD}$, the absence of coordination costs results in a relatively higher SW. This suggests that for higher values of $\alpha$, the inefficiencies and costs associated with revenue-sharing coordination reduce its effectiveness in enhancing SW compared to the status quo. What cannot be ignored is that the positive effect of revenue sharing on SW appears when $\alpha$ is very low.

6. Concluding Remarks

6.1. Discussion

6.1.1. Theoretical Contributions

This paper studies a SC comprising a supplier, a TPR, and a retailer. The retailer places orders of genuine products and remanufactured products with the supplier and the TPR, respectively (i.e., Model O). In practice, the genuine products and the remanufactured products encounter challenges associated with demand uncertainty. Meanwhile, the retailer holds an information advantage over the supplier and the TPR owing to abundant first-hand sales data; thus, information asymmetries occur. For the retailer, there is a profit margin in taking advantage of this information gap. Therefore, the retailer may resort to a risky, unfair competition strategy—relabeling a portion of remanufactured products as genuine products to pursue heightened profits, constituting brand deception (i.e., Model BD). To avoid the retailer’s deception practice, the supplier and the TPR may adopt AIT to obtain quality information (i.e., Model BA) or the supplier may offer a revenue-sharing contract to the retailer (i.e., Model C). This paper focuses on the trade-offs between brand reputational damage, industry adaptation resistance of AIT, demand expansion, and revenue-sharing coordination in SC members’ preferences. In addition, this paper also studies the total profit, CS, and SW to show how brand deception’s social impact can be inhibited under different models.

6.1.2. Practical Implications

Our findings provide insightful implications in practice, including the following:

(1) Technology vendors and industry leaders such as BYD, Bosch, and Weichai should focus on reducing BCT adoption costs through innovation or subsidies, encouraging widespread adoption among retailers pivotal in combating brand deception in the SC. Small and medium-sized enterprises such as Miracle Automation Engineering and CARLOS should carefully evaluate the cost versus benefit of BCT implementation. While BCT can significantly boost consumer trust and demand, its cost-effectiveness must be ensured. (2) Mercedes-Benz could embed RFID tags or QR codes on its spare parts and combine them with blockchain technology to develop a traceability platform, allowing consumers to easily verify the authenticity of each accessory. In cooperation with authorized dealers, Mercedes-Benz could design a profit-sharing model based on sales volume and service quality. In addition to profit sharing based on sales quantities, it could also obtain additional rewards by providing high-quality after-sales service and customer experience. (3) Taking Louis Vuitton in the luxury goods industry as an example, LV could avoid brand deception by implementing stricter pricing and product line strategies. Specifically, LV could launch special second-hand series, such as “Certified Pre-Owned” or “Pre-Loved” series, clearly marking these products as “refurbished” or “remanufactured”, so that consumers can clearly understand the true situation of the products when purchasing, thereby reducing misleading labeling and brand deception. In addition, LV could label these refurbished products as “limited series” and price them at a lower discount, while marking them as “officially certified refurbished”. All such products could be sold through the brand’s official website, specialty stores, or authorized e-commerce platforms to ensure that their quality and authenticity are guaranteed by the brand. In this way, LV could not only maintain its high-end brand image, but also meet the market’s demand for low-priced luxury goods while avoiding the potential risks of brand deception. (4) When BCT and revenue-sharing coordination were introduced, retailers’ profit margins for brand deception were greatly compressed. However, being suppressed does not mean being completely absent. Subsequently, we believe that a new type of brand deception will emerge, and TPRs may collaborate with retailers to carry out brand deception and form a tacit understanding to jointly deceive suppliers. How to prevent this kind of deception will be a new challenge for suppliers. (5) Enhancing transparency through BCT or improving cooperation through revenue sharing directly benefits consumer trust and CS. Retailers should abandon short-term deception benefits and focus on reducing deception not just to comply with regulations but to significantly boost consumer perception, market share, and SW.

Our findings also provide specific action guidance in

Table 4. Specific action guidance.

Suppliers are advised to (i) Adopt AIT when the relabeling rate is relatively low and the industry adaptation resistance falls within the valid range of the bi-interval adaptation effect; (ii) Adopt revenue sharing when the relabeling rate is high and reputational damage is low; (iii) Eliminate brand deception when reputational damage is high (i.e., fashion sector).

TPRs are advised to (iv) Adopt AIT when industry adaptation resistance falls within the valid range of the bi-interval adaptation effect; (v) Avoid revenue sharing; (vi) Eliminate brand deception when the relabeling rate is high.

Retailers are advised to (i) Stop brand deception when the relabeling rate exceeds the threshold; (ii) Adopt AIT when industry adaptation resistance is low (i.e., pharma, fashion sectors); (iii) Accept revenue sharing when the relabeling rate is low and reputational damage is high (i.e., auto sector).

Retailers benefit from brand deception in the case of (i) A sufficient low relabeling rate; (ii) Strictly required region of relabeling rate, reputational damage, and industry adaptation resistance.

Triple win in the case of (i) Reputational damage, industry adaptation resistance, and relabeling rate being low; (ii) Both industry adaptation resistance and relabeling rate being high enough.

:

6.2. Conclusions

The main findings of this paper are as follows:

(1) Brand deception negatively affects the supplier’s profits while both the TPR and retailer benefit at low deception levels, with the TPR being the biggest beneficiary. Brand deception results in decreased (increased) sales of genuine (remanufactured) products; thus, the supplier (TPR) tends to set a lower (higher) wholesale price. The asymmetric profit erosion effect was identified to highlight the dynamic process by which brand deception leads to the gradual erosion of profits for the supplier, retailer, and TPR. (2) For both supplier and TPR, AIT adoption results in higher wholesale prices than in Model BD. In addition, order quantities surpass those in Model BD, provided that industry adaptation resistance remains below a certain threshold. The bi-interval adaptation effect indicates that AIT is particularly effective in industries with low adaptation resistance. When both the relabeling rate and industry adaptation resistance are low (high), Model BA (Model O) achieves a triple win. (3) Regarding social impact, brand deception can lead to a temporary increase in total profits and SW compared to Model O. However, once brand deception occurs, CS declines. Sequentially, when the industry adaptation resistance is low, AIT can significantly improve the total profits, CS, and SW. Compared to Model BD, revenue sharing offers slight advantages in CS but significant disadvantages in SW. We indicate two possible research directions. First, we exclude the quality value of the remanufactured product which might impact the yield quantity and wholesale price of the remanufacturer. As a response, the retailer’s preference might be changed and even start to self-band on recycling products and directly remanufacturing for resale. Second, the study excludes the impact of external factors such as government-imposed environmental carbon taxes on genuine products and subsidies for launching remanufactured products. Future research could investigate how these external environmental policies affect the market’s valuation of both genuine and remanufactured products. Carbon taxes might reduce the demand for genuine products, while subsidies could significantly increase the appeal of remanufactured products, potentially reshaping the competitive landscape between genuine and remanufactured products.