DeepSeek AI in action: exploring adoption dynamics with UTAUT, task-technology fit and initial trust models

Purpose DeepSeek is a new artificial intelligence (AI) landscape with significant benefits. However, users accustomed to OpenAI may resist the transition. In this rapidly evolving AI landscape, understanding the factors that drive the adoption of DeepSeek is crucial for both developers and users. This study aims to examine the adoption dynamics of DeepSeek AI using the Unified Theory of Acceptance and Use of Technology (UTAUT-3), Task–Technology Fit (TTF) and the Initial Trust Model (ITM). Design/methodology/approach Data was gathered from 621 Generation Z respondents in Northern India through an online survey. Nonprobability sampling was applied for convenience, using a contact snowball technique. Findings Partial least squares structural equation modeling analysis revealed that Performance Expectancy, Effort Expectancy, Social Influence Facilitating Conditions and Hedonic Motivation significantly influence Behavioral Intention to adopt DeepSeek AI. Performance Expectancy has the strongest impact. Effort Expectancy, which relates to ease of use, has a lesser effect. TTF findings indicate that DeepSeek AI’s functionalities, such as real-time responses and data security, align with user tasks, enhancing perceived usefulness but not directly driving adoption. Instead, trust and social influence play essential roles. Initial Trust influences adoption through Behavioral Intention, moderated by Structural Assurances and Personal Propensity Trust, emphasizing the need for transparency and reliability. Research limitations/implications A notable limitation of this study is its focus on a single AI Platform, DeepSeek AI; therefore, the generalizability of the findings to other AI tools or platforms cannot be assured. Moreover, the analysis is based on self-reported data from users, which might bear biases, including social desirability and response bias. To overcome these limitations, future studies should examine a wider range of AI technologies across different domains and move beyond self-reported usage behaviors by using more objective measures, such as system logs or actual usage data. Practical implications Organizations developing AI tools such as DeepSeek AI should enhance performance and ease of use by aligning system capabilities with user tasks. Trust plays a central role in adoption; therefore, transparent policies, explainable AI and strong ethical standards are essential. Social influence, supportive conditions and hedonic motivation further strengthen behavioral intention and engagement. Ultimately, trust shapes the translation of intention into actual usage, emphasizing the need for transparent and user-centric AI solutions. Originality/value This study uniquely integrates UTAUT-3, TTF, and the ITM to provide a multidimensional understanding of DeepSeek AI adoption among Generation Z users in Northern India. The findings highlight the critical roles of performance expectancy, social influence and trust, while also revealing the limited influence of habit on adoption behavior. This research offers fresh insights into how emerging users engage with advanced AI platforms. This study advances current literature by emphasizing the moderating effect of trust and the need for user-centric design, transparency and ethical AI deployment to ensure sustainable adoption and long-term engagement.