​Model conduct, Pg6

Key Highlights:

• India currently regulates Artificial Intelligence (AI) through the IT Act, financial sector regulations, and data protection rules.
• China has proposed new rules targeting emotionally interactive AI services, requiring warnings against excessive use and intervention for extreme emotional states.
• India's approach is less intrusive but lacks a specific duty of care regarding AI product safety, particularly for psychological harms.
• The Ministry of Electronics and Information Technology (MeitY) uses IT Rules to address deepfakes and requires labeling of "synthetically generated" content.

Detailed Insights:

• India's regulatory approach to AI is primarily reactive, with MeitY responding to issues like deepfakes rather than proactively establishing safety standards.
• Financial regulators like the RBI and SEBI are taking preemptive steps by setting expectations for model risk management and accountability in AI tool usage.
• India lags behind the U.S. and China in developing its own frontier AI models, making it crucial to avoid over-regulation that could stifle domestic innovation.
• A practical approach involves nurturing domestic AI capabilities through improved access to resources and strategic public procurement.
• India should focus on regulating the downstream use of AI in high-risk contexts, such as requiring incident reports from companies deploying AI products.

Key Concepts Involved:

• Artificial Intelligence (AI): Simulation of human intelligence processes by computer systems.
• Deepfakes: Synthetic media in which a person in an existing image or video is replaced with someone else's likeness.
• Duty of Care: A legal obligation to avoid acts or omissions that could reasonably be foreseen to cause harm to others.