Page 170 - ISC PROCEEDINGS 21.4

into their personal space. This finding aligns with the study by Lucia-Palacios and Pérez-
López (2021), which suggests that higher levels of autonomy in intelligent systems may
lead users to feel a loss of control and greater concern about technological interference in
their personal lives. However, perceived autonomy does not have a significant effect on
privacy disclosure risk. This result suggests that in the context of AI-driven virtual
streamers, the traditional relationship between autonomy and privacy risk may have
shifted. According to Menard and Bott (2025), AI systems are capable of combining
fragmented behavioral data to infer personal information that users have never explicitly
disclosed. In such cases, the risk lies not only in the act of sharing information but also in
the algorithm’s ability to analyze and reconstruct data, making users’ perceived control
over input data less relevant in shaping privacy disclosure risk. In addition, the
phenomenon of digital resignation (Choung et al., 2024) and decision fatigue related to
privacy management (Herriger et al., 2025) may lead users to become indifferent toward
data control mechanisms, thereby weakening the relationship between perceived
autonomy and privacy risk.
Finally, the results indicate that both intrusiveness risk and privacy disclosure risk
have positive effects on resistance intention. This finding highlights the central role of risk
evaluations in shaping users’ behavioral responses. The results are consistent with Aw et
al. (2024), who found that when users perceive risks related to personal privacy or
technological intrusion, they tend to resist such technologies. Similarly, Shao and Ho
(2025) demonstrated that when users become more aware of the potential negative
consequences of virtual streamers, they are more likely to translate these evaluations
into behavioral intentions, such as reducing interaction or refusing to use the service.
4. Conclusion
This study investigates how users evaluate and respond to virtual streamers in the
livestream commerce sector. By examining the psychological factors that either
encourage or hinder acceptance, the study provides a clear picture of the modern digital
shopping experience. The findings suggest that user acceptance is not a simple reaction to
new technology but rather a careful calculation of risks and interaction quality. The
results indicate that user resistance is largely driven by cognitive evaluations of risk. In
particular, perceived intrusiveness and privacy disclosure risks are the primary reasons
why users remain hesitant. When users feel that their personal space is being invaded or
that their data security is at stake, they are much less likely to engage with virtual selling
environments. These concerns represent the most significant barriers that developers and
platforms must address to build long term user commitment.
At the same time, the study highlights how highly advanced interaction features can
inadvertently raise barriers. The findings demonstrate that excessive perceived
anthropomorphism and autonomy trigger feelings of intrusiveness and ontological
ambiguity. When a virtual streamer becomes too human-like or acts too independently,
users feel a loss of control and boundary invasion, which in turn fuels their resistance.
Although autonomy does not directly translate into privacy disclosure fears, its impact on
perceived intrusiveness is significant. Furthermore, this study offers insights specific to
the market in Vietnam, where livestream commerce is growing rapidly. In an environment
where digital regulations and user awareness are still evolving, concerns about trust and
data security are particularly influential. The results suggest that while users appreciate
the efficiency of AI, their willingness to adopt these tools depends heavily on how well
platforms protect user privacy and manage digital boundaries.

169