Article
AI is rewriting the threat map for India's BFSI sector
CERT-In data shows cyberattacks on Indian banks have more than doubled as attackers turn to autonomous AI tools.
17 Jul 20261 min readAI4U Desk
Cyberattacks on India's banks more than doubled from 1.4 million in 2021 to 2.9 million in 2025, according to CERT-In's Digital Threat Report 2025-26. MeitY's report highlights a growing threat of AI asymmetry to the BFSI sector, where hackers leverage artificial intelligence to launch multi-vector attacks simultaneously against banks, cloud systems, and vendors.
This shift in capability was starkly visible in November 2025, when a Chinese-linked group used Anthropic's Claude to run ninety percent of a global espionage operation attacking thirty firms, including financial institutions. Furthermore, frontier AI models have successfully reproduced over half of 405 historical cyberattacks against blockchain software in tests, resulting in $550 million in simulated losses.
Attack velocity is now severely outpacing traditional bank defenses. AI shrinks discovery-to-exploit timelines to machine speed, while compliance gates still move at human speed. Malware has effectively become autonomous.
To counter these threats, CERT-In's report prescribes three operational fixes that India's banks and fintechs must adopt. Defenders require AI-speed operations rather than sequential incident response, which remains structurally unprepared for parallel, AI-run attacks on the BFSI sector.
