PERSPECTIVES FROM SELECT INDIAN IT COMPANIES ON AI GOVERNANCE: EVIDENCE FROM A SURVEY AND EXPERT INTERVIEWS

Abstract

Artificial intelligence (AI) is moving into routine business use, but many Indian IT companies still find it difficult to translate Responsible AI principles into everyday governance. This paper examines the drivers of Responsible AI Governance in Indian IT companies using a structured survey of 177 practitioners and a follow-up expert interview with some senior security and AI governance leaders to contextualise the findings. The survey instrument was built around six constructs: Transparency and Ethics in AI Systems (TEAS), Accountability and Ethical Leadership (AMEL), Organisational Culture and Ethical Climate (OCEC), Regulatory and Legal Compliance (RLC), Data Governance and Quality Control (DGQC), and Responsible AI Governance Effectiveness (RAGE). After cleaning, 176 complete cases were used for multivariate testing. Reliability was satisfactory across all constructs, with Cronbach's alpha ranging from 0.71 to 0.88. Descriptive results show that respondents rated transparency and data governance relatively positively, while regulatory compliance received the lowest mean score. Correlation analysis found that all five governance dimensions were positively associated with RAGE, with RLC and OCEC showing the strongest relationships. Multiple regression explained 74.3 per cent of the variance in RAGE. In the presence of the other predictors, RLC and OCEC remained statistically significant, while TEAS, AMEL and DGQC did not retain independent significance. The interview adds practical context: the participant emphasised centralised governance, approved toolsets, mandatory checklists, training, and leadership support as the basis of responsible AI practice. The findings suggest that Responsible AI Governance in Indian IT companies is not built by policy statements alone. It grows where regulatory readiness and ethical culture are taken seriously and where governance is treated as an organisational capability rather than a compliance exercise.