Indian govt plans AI oversight board, mandates training for bureaucrats

Representative image

The Indian government is ramping up oversight of artificial intelligence (AI) in governance, mandating AI training for bureaucrats and proposing a dedicated board to regulate government use of the technology.

Under the recently released AI Competency Framework for Public Officials—part of the broader IndiaAI Mission—the government has recommended forming an AI governance board to evaluate, approve, and monitor AI applications. The board would ensure compliance with ethical and regulatory standards.

"Establish a dedicated AI Governance Board with the objective of reviewing and authorising AI applications," the framework says.

The framework does not explicitly state whether the governance board will review and authorise AI applications in the private sector, or if its mandate is strictly limited to government-led AI initiatives.

The framework also emphasises that government officers must develop a critical understanding of AI’s ethical and security challenges.

"The governance board should ensure that AI projects adhere to established legal and ethical guidelines, both domestic and international. It will also play a key role in reviewing AI initiatives throughout their lifecycle," the framework outlines.

Beyond regulatory approval, the panel is expected to offer guidance to ensure AI applications are both technically sound and socially responsible.

Last year, a controversy erupted after the government, in an advisory, asked artificial intelligence platforms to seek its permission before launching an AI product in the country. The advisory was rolled back after it faced huge criticism from the industry who termed it as 'regressive'  and capable of "stifling innovation".

Alongside governance reforms, the framework requires officials to assess AI risks, biases, and unintended consequences before implementation.

This marks a shift from a tech-first approach to a risk-aware governance model, requiring policymakers to scrutinise AI-driven decisions for fairness, transparency, and potential harm.

A critical issue raised in the framework is bias in AI systems, particularly concerning the representation of marginalised communities.

"Marginalised groups are often underrepresented or misrepresented in training data. This can result in flawed decision-making with real-world consequences," the document states.

To counteract this, it mandates rigorous audits of AI models against baseline datasets and encourages the inclusion of independent assessments to gauge ethical and performance standards.

The three-tier competency framework structures AI training based on roles and responsibilities. Level 1 is for policy-makers setting AI governance strategies, Level 2 covers mid-level officials overseeing AI-driven programs, and Level 3 focuses on implementation teams handling day-to-day AI operations.

To ensure ethical AI deployment, the framework requires comprehensive documentation of AI models, including transparency on algorithm design and data sources. Additionally, it calls for human oversight mechanisms wherever AI affects public services or fundamental rights.

With AI expected to add $450–500 billion to India’s GDP by 2025, the government pushing for AI adoption but also working to institutionalise safeguards that enhance accountability, transparency, and fairness.