<p>Psychological safety has emerged as an important dimension of organisational culture because employees rarely contribute their knowledge, concerns or unconventional ideas when interpersonal risk is high. Although organisations frequently encourage innovation, many employees remain silent because they anticipate criticism, humiliation, loss of status, unfavourable performance evaluation or damage to their careers. Such silence restricts early identification of errors, prevents circulation of tacit knowledge and weakens organisational learning. The present paper examines psychological-safety culture as an antecedent of employee voice and innovative work behaviour, with knowledge-sharing behaviour functioning as a mediating mechanism. The study integrates psychological-safety theory, social-exchange theory, organisational-learning theory and the componential theory of creativity. A quantitative, explanatory and cross-sectional research design is proposed for employees working in knowledge-intensive organisations, including technology companies, universities, hospitals, startups and research institutions. The conceptual model proposes that psychological-safety culture directly encourages employee voice, knowledge sharing and innovative work behaviour. It further proposes that knowledge sharing mediates the effect of psychological safety on employee voice and innovation, while employee voice provides an additional behavioural pathway through which shared knowledge is converted into workplace innovation. A stratified sample of approximately 400 employees is recommended, with data collected through a structured five-point Likert-scale questionnaire. Confirmatory factor analysis and structural equation modelling are proposed for assessing the measurement and structural models.</p>

<p>The paper concludes that organisations can improve innovation by developing inclusive leadership, fair responses to failure, accessible voice mechanisms, knowledge-sharing routines and credible follow-up procedures. Since no primary dataset has been used, the paper presents anticipated relationships and an empirical testing framework rather than fabricated statistical findings.</p>