A senior Indian government official has stressed that public institutions and government departments should exercise caution before widely adopting artificial intelligence (AI) solutions, saying they should not rush into use until the technologies are thoroughly tested and proven reliable. The comments were made during the India AI Impact Summit 2026 in New Delhi, where discussions focused on how best to integrate AI into public systems and policy.
Rohit Bhardwaj — Deputy Director General in the Data Informatics & Innovation Division of the Ministry of Statistics and Programme Implementation (MoSPI) — emphasised that governments and institutions need to ensure that AI tools are trustworthy before deployment, warning against being “gung-ho” about untested systems. He highlighted that AI can interpret the same data in different ways even with similar prompts, underscoring the potential for inconsistent outputs if tools are not evaluated carefully.
Bhardwaj also pointed out that preparing data for AI use is a critical step itself, urging departments to make their information AI-ready by converting it into machine-readable formats with clear context files, semantics, and metadata rather than unstructured PDFs. Without this groundwork, he said, even well-designed AI systems might produce unreliable or misleading results.
The caution aligns with broader conversations at the summit about responsible AI adoption — balancing innovation with safety and trust. The official said he would personally adopt AI tools in his work only if they meet reliability criteria, and he underscored the importance of raising awareness about what AI readiness entails among government ministries and agencies.
