Font size:
Print
AI Regulation Report
Context:
Last week, the Ministry of Electronics and Information Technology (MeitY) released its Report on the Regulatory Framework for AI in India.
More on News
- While the recommendations in the report are largely pragmatic and non-controversial, the underlying principles raise questions about whether the approach will hinder India’s AI industry from reaching its full potential.
Overreliance on Global North Principles
- A significant critique of the report lies in its heavy reliance on regulatory frameworks developed by countries in the Global North.
- While studying these approaches can be valuable, adopting them without adapting to India’s unique socio-economic context may lead to challenges.
- For developed nations, AI is often viewed as a tool to enhance the efficiency of already well-functioning systems.
- As such, they can afford to prioritise mitigating risks over reaping benefits.
- In contrast, India sees AI as a transformative necessity—a critical tool to bridge gaps and deliver benefits to underserved and marginalised communities.
- For India, overly cautious regulation might stifle innovation and prevent the country from leveraging AI’s full potential.
Techno-Legal Measures: A Misguided Focus?
- The report places considerable emphasis on techno-legal solutions for regulation, but this approach may not always be appropriate.
- For instance, the suggestion to use the MeitY consent artifact to determine liability for AI-related harms is problematic.
- The report proposes assigning immutable, unique identities to participants in the AI ecosystem, tracked using the consent artifact.
- However, this approach conflicts with the privacy-by-design principles of the Data Empowerment and Protection Architecture (DEPA).
- Consent managers, by design, cannot access the content being transferred or establish unique identities.
- Modifying the system for such purposes would undermine its core principles and practical functionality.
Lack of Actionable Solutions for Copyright Concerns
- The report also discusses copyright challenges linked to AI development and deployment.
- However, instead of providing clear solutions, it calls for yet another round of consultations.
- Amending the Indian Copyright Act to include fair use provisions and text-and-data mining exemptions—similar to those in other countries—would have been a practical recommendation.
- The absence of such actionable suggestions leaves developers without the clarity they need to innovate confidently.
Positive Recommendations to Build On
Despite these concerns, the report includes several commendable recommendations:
- No AI-Specific Legislation: The report rightly concludes that existing laws and regulations are sufficient to address the risks posed by AI, avoiding the need for specific AI legislation.
- Regulatory Training: It emphasises the need to train regulators and law enforcement agencies on the potential harms of AI, enabling them to adapt their oversight strategies effectively.
Institutional Recommendations
The report proposes two significant measures:
- An Inter-Ministerial AI Coordination Committee: This would oversee AI governance across sectors, addressing the cross-cutting nature of AI-related challenges.
- A Technical Advisory Secretariat: Tasked with providing technical advice and evaluating AI incidents, this body would ensure swift and informed decision-making.
A Balancing Act
- If judged solely on its recommendations, the report offers a balanced mix of agile governance and light-touch supervision.
- However, the thinking that underpins these recommendations suggests a Western-centric, risk-averse approach.
- If this mindset extends to enforcement, India’s AI companies may find it difficult to innovate in a highly competitive global landscape.
India’s unique developmental challenges demand a regulatory framework that prioritises AI’s transformative potential over excessive caution. Striking the right balance between risk and opportunity will determine whether this framework becomes a catalyst for progress or a barrier to innovation.
