Feeling empowered by AI chatbots was linked to stronger misuse intentions

A study of 736 users found that chatbots’ capability and accessibility were associated with greater feelings of autonomy and competence, which were in turn linked to intentions to misuse them.

Editorial illustration — not from the study.

The study examined whether features of generative AI chatbots, particularly their capability and accessibility, were associated with users’ feelings of autonomy and competence. It then assessed whether those feelings were linked to intentions to use the systems in ethically questionable ways.

The researchers found that chatbot capability and accessibility were positively associated with both autonomy and competence. These two feelings were, in turn, positively associated with misuse intention. Situational anxiety changed only the relationship between competence and misuse intention, while explanations based on moral rationalization showed mixed associations.

What users reported

Among the 736 generative AI users studied, perceived chatbot capability and accessibility were positively associated with feelings of autonomy and competence. Greater autonomy and competence were positively associated with intentions to misuse generative AI. Situational anxiety moderated only the link between competence and misuse intention. Measures of neutralization—ways people may justify questionable behavior—showed mixed associations, suggesting that empowerment-related perceptions may be more central than conventional rationalization in explaining the reported misuse intentions.

Why empowerment matters

The findings point to a possible unintended ethical consequence of making AI chatbots feel more capable and accessible: users may also feel more empowered in ways associated with questionable use. The study suggests that efforts to understand responsible AI use should consider users’ feelings of autonomy and competence, not only whether they can rationalize their behavior after the fact.

Evidence and limits

The researchers analyzed survey data from 736 generative AI users using structural equation modeling, a statistical method for examining relationships among several measured factors. Because the study was cross-sectional—capturing responses at one point in time—it cannot show that chatbot capability or feelings of empowerment cause misuse intentions. The findings show associations between these factors.

// Source

Asia Pacific Journal of Marketing and Logistics · 2026 · DOI: 10.1108/apjml-04-2026-0774

View paper (DOI)Open access versionPublished 2026-09-21

Authors: Shiu‐Wan Hung, Jyun-Hao Jian

Institutions: National Central University

Programmable hardware speeds matrix calculations up to 6.3 times

Researchers tested a programmable hardware design that imitates selected quantum-computing ideas to run matrix calculations in parallel. In representative benchmarks, the design reached up to 6.3 times the speed and 4.1 times the energy efficiency of conventional matrix co-processors and embedded graphics processors.

2 days ago·13 views

AI & Computing

AI and docking agreed in some drug screens, but not all

Researchers evaluated three AI systems on hundreds of molecules bound to a SARS-CoV-2 protein and in three prospective drug screens. The results suggest that combining AI-based predictions with traditional docking may help prioritize candidates, although no single approach worked best in every test.

4 days ago·50 views

AI & Computing

A hardware monitor detected simulated attacks on car networks in 14 microseconds

Researchers developed a hardware-assisted system intended to detect message injection, spoofing, flooding and impersonation attacks on automotive networks. In tests using simulated traffic and a prototype automotive processor system, it reported an average detection time of 14 microseconds and less than 2 percent CPU overhead.

5 days ago·43 views