Extended Technology Acceptance of Workplace GenAI Copilots: Evidence from PLS-SEM

Organizations are rapidly introducing generative-AI copilots to augment knowledge work, yet clarity is still needed on what is a significant predictor of employee adoption. This study proposes an Extended Technology Acceptance Model tailored to the copilot context that integrates algorithmic transparency, trust, perceived risk, social influence, and facilitating conditions alongside the classic components of perceived ease of use, perceived usefulness, attitude, intention, and use. Survey data from employees in the Gulf region were analyzed with variance-based structural equation modeling. The reflective measurement model exhibited strong loadings, reliability, and discriminant validity. Results show that perceived ease of use enhances both usefulness and attitude, while usefulness, attitude, and social influence together elevate behavioral intention. Transparency operates as an upstream design lever: it increases trust and reduces perceived risk, and trust in turn strengthens perceived usefulness. Intention and facilitating conditions jointly translate into actual use of the copilot. Direct paths from risk to intention and from trust to intention were weak once transparency and usefulness were considered, indicating primarily indirect influence. Predictive assessments support the practical relevance of the model. The findings suggest clear managerial levers includes invest in transparent explanations, nurture trust, build champions to shape norms, and ensure robust support and accessso organizations can convert enthusiasm about AI into sustained, responsible use.