India Takes Action: SEBI Issues Directive on AI Model Cyber Risks

The state machine has processed the "threat - reaction" cycle in a week and a half. On May 4, 2026, the Securities and Exchange Board of India (SEBI) announced the issuance of an official advisory act on the risks of generative AI in the financial sector.

This document is the direct execution of the Indian Ministry of Finance's mandate from April 25. India is officially falling in line with the Asia-Pacific countries (Australia, Singapore, Japan) building regulatory defenses against algorithms. The trigger was the elite Anthropic Mythos model, capable of finding 0-day vulnerabilities in banking code. The SEBI regulator is making it clear to brokers and exchanges: the use of "frontier" neural networks requires the implementation of the strictest audit systems. Financial market cybersecurity now officially accounts for autonomous algorithms as a systemic attack vector.

Source: SEBI / Reuters / Business Standard