India’s AI Governance Guidelines Report – FAQs answered

In February 2025, the Prime Minister of India, Narendra Modi, co-chaired the AI Action Summit in Paris along with the President of France, Emmanuel Macron. Underlining the need for transparency and bias-free datasets, Modi proclaimed that India is leading in AI adoption and that the world was “at the dawn of the AI age that will shape the course of humanity”1. He emphasised how the adoption of AI would not lead to job losses and that there was a need to skill and reskill people for an AI-driven future.  
 
About a month before the AI Action Summit in January 2025, the Ministry for Electronics and Information Technology (MeitY) in India released a 22-page report on AI governance guidelines development (“Report”)2. The Report was drafted by a Sub-Committee on “AI Governance and Guidelines Development” constituted under the guidance of the Advisory Group, chaired by the Principal Scientific Advisor. The public consultation on the Report concluded on 27 February 2025.  
 
The Report outlines MeitY’s approach to responsible AI development and deployment.

Does India truly lead in AI adoption? Does the Report address issues pertaining to job losses caused by AI or lack of transparency and fairness in AI systems? The following blog article answers a few of these FAQs.
 

1. What has been the state of the Indian AI legislative framework till now?: India has yet to introduce a dedicated AI law but has relied on existing legal frameworks and sectoral guidelines (RBI, TRAI, etc.) to govern AI applications. Prior to the AI Governance Guidelines Report, India’s approach to AI regulation was fragmented, sector-specific, and largely focused on principles rather than enforceable laws. There exists a draft proposal for a new Digital India Act, and this includes a reference to “regulation of high-risk AI systems”3. Additionally, there are advisories that have been published and withdrawn, and then republished, regarding obtaining permission for the deployment of certain AI models or for regulating the distribution of deepfakes.4 However, the approach as of now has been that of “light-touch regulation” with a focus on self-regulation, ethical AI use, and voluntary compliance rather than legally binding frameworks

2. What is the AI Governance Guidelines Report? : The AI Governance Guidelines Report is a policy framework released by MeitY through a sub-committee working on AI governance with the aim of regulating and guiding responsible development, deployment, and use of artificial intelligence in India. It is the outcome of a gap analysis undertaken by the sub-committee whereby it has taken three aspects into consideration: i) The need to enable effective compliance and enforcement of existing laws, ii) The need for transparency and responsibility across the AI ecosystem in India, and iii) The need for a whole-of-government approach (an approach which intends to reduce inefficiency and gaps in AI regulation due to fragmentation in governmental departments).

3. What are the key objectives of the Report?: The primary objectives of the Report are:

• • Ensuring ethical AI development. 
  • Promoting transparency and accountability.  
  • Mitigating bias and discrimination in AI systems. 
  • Protecting privacy and data security. 
  • Encouraging innovation while addressing risks

4. Does the Report introduce new AI regulations?: No, the Report does not introduce any legally binding regulations. It outlines a set of voluntary guidelines and best practices aimed at encouraging responsible AI development. These recommendations serve as a framework for ethical AI adoption but do not impose mandatory compliance or penalties for violations. The approach relies on self-regulation by companies and institutions rather than government-enforced legal obligations. It does, however, recommend the formation of an “Inter-Ministerial AI Coordination Committee or Governance Group” to coordinate efforts between key institutions for the implementation of a “whole-of-government” approach. Additionally, the Technical Secretariat, a technical advisory body and coordination focal point, is to be established for the Group, which may be staffed by existing MeitY officials as well as lateral hires. It may also be tasked with examining the viability of technical solutions to address AI-related risks.

5. What are the key principles outlined in the Report?: Aligning itself with the efforts of OECD, NITI and NASSCOM, the Report emphasises the following principles: 

• • Fairness and non-discrimination.
  • Safety, reliability and robustness.
  • Transparency and explainability.
  • Accountability.
  • Privacy and security.
  • Human-centred values and “do no harm”.
  • Inclusive and sustainable innovation.
  • Digital by design governance.

6. How does the Report address bias and discrimination in AI?: The Report highlights the need to identify and mitigate biases in AI systems to prevent discrimination against marginalised communities. It recommends the development of an “AI incidence reporting database” for the systematic collection of evidence to facilitate government initiatives. This database may be aligned with the OECD AI Incidents Monitor.5 There are no mentions of conducting bias audits at various stages of AI development, ensuring the use of diverse and representative datasets, and implementing regular impact assessments to detect and address discriminatory outcomes.

7. What does the Report say about AI and privacy? : The Report underscores the importance of privacy-by-design and security-by-design, meaning AI systems should be built with data protection measures from the outset rather than as an afterthought. It aligns with India’s Digital Personal Data Protection Act, 2023, requiring AI developers to ensure lawful data processing, user consent, and secure storage of personal data. However, it does not provide clear enforcement mechanisms or address concerns about AI-driven collection of data or transparency surrounding the same.

8. Is there any mention of AI in law enforcement and surveillance?: Yes, the report recognises AI’s potential in public safety and security, particularly for law enforcement applications. In view of the same, it states that the government may adopt the recommendations in the Report, including the transparency and governance measures. Despite these concerns, the Report does not explicitly state anything with regard to surveillance or ban controversial AI-powered surveillance technologies or provide strict oversight mechanisms for their use.
 

9. What does the Report propose for AI governance in the private sector?: The Report encourages companies to adopt ethical AI practices, especially through voluntary reports and disclosures. It also recommends conducting risk assessments to identify potential harms and establishing grievance redressal mechanisms to address AI-related discrimination or errors in collaboration with the Governance Group. However, compliance with these measures is voluntary, and companies are not legally bound to implement them. This raises concerns about corporate accountability in AI governance.
 

10. Does the Report cover AI’s impact on jobs and labour rights?: The Report lacks detailed provisions on labour protections, such as safeguards against AI-driven exploitation, unfair job automation, or algorithmic discrimination in hiring. Without stronger worker protections, AI could worsen inequalities in India’s labour market rather than fostering inclusive growth.
 

11. What are the Report’s recommendations for AI safety and security?: The Report suggests multiple measures to ensure AI safety. These include implementing AI incident reporting mechanisms to track failures and risks and strengthening cybersecurity measures to prevent AI-related cyber threats. These recommendations, unfortunately, lack clear regulatory enforcement for ensuring compliance. However, in what could be deemed as a positive, the Report notes that for AI systems providers or deployers, safe harbour provision (liability for third-party content) would not be applicable.
 

12. How does the Report approach AI in healthcare, education and governance?: The Report encourages AI innovation in healthcare (diagnostics, patient care), education (personalised learning), and governance (public service automation) while emphasising the need for accuracy checks and data protection mechanisms. However, it does not specify legal safeguards against AI-driven biases, particularly in areas like automated decision-making for welfare benefits or predictive policing. This can lead to situations where, without stricter oversight, AI may end up reinforcing existing societal inequalities rather than solving them.
 

13. What role does environmental sustainability play in AI governance?: The Report fails to acknowledge the high energy consumption of AI systems, particularly in a country like India, which struggles with the allocation of already insufficient resources for a large population. There are no mandatory sustainability guidelines. Companies can, thus, continue deploying resource-intensive AI models without environmental accountability.
 

14. How does India’s AI governance approach compare to the EU AI Act?: India’s AI guidelines are non-binding and driven by industry. This entails a situation where compliance is voluntary, whereas the EU AI Act enforces strict legal obligations based on a risk-based regulatory model. The EU classifies AI applications into prohibited, high-risk, and minimal-risk categories, with strict penalties for non-compliance. In contrast, India’s approach relies on self-regulation. This raises concerns regarding the lack of enforcement, weak corporate accountability, and insufficient protection against AI-related harms.
 

15. Does India’s approach to AI regulation affect EU businesses in any manner?: The absence of AI regulation in India creates regulatory uncertainty for EU businesses operating in or outsourcing to India, particularly under the EU AI Act and the GDPR. Since India lacks binding AI laws, compliance mismatches may arise, making it difficult for EU companies to ensure that Indian AI vendors meet European transparency, accountability, and bias mitigation standards.

Additionally, the lack of mandatory risk-based classification in India contrasts with the EU AI Act, which categorises AI systems based on their risk level. This divergence means EU businesses must independently assess AI risks when outsourcing AI solutions to Indian firms, adding compliance costs.