JP Mishra, Founder and CEO of Deep Algorithm, said cyberattacks are no longer primarily human-led, but driven by intelligent systems that evolve in real time. “It is like a war of AI versus AI. Conventional security systems were built to tackle attacks coming from human attackers. But nowadays, attacks are not from humans; it is where AI agents themselves are carrying out attacks,” he said in an interview with CNBC-TV18.
These AI agents, Mishra noted, operate at speeds and levels of adaptability that legacy systems cannot match, continuously learning from each attempted breach and modifying attack vectors. In response, organizations are being forced to deploy equally sophisticated defenses. “To tackle these kinds of attacks, we require AI defense systems, where autonomous AI red teams continuously scan and find vulnerabilities, and autonomous AI blue teams fix and patch them in real time,” he said.
The implications are particularly significant for BFSI players, where security breaches can have systemic consequences. As financial institutions deepen their reliance on digital infrastructure—from payments to lending and wealth management—the attack surface has expanded, making traditional signature-based tools increasingly ineffective.
From an investor standpoint, this technological shift is driving a clear reallocation of capital. Bhaskar Majumdar, Managing Partner at Unicorn India Ventures, said the funding is moving decisively towards AI-native platforms that can continuously learn and adapt. “There is a shift in capital allocation—clearly away from static tools, which were more signature-based and rules-driven, towards adaptive, AI-native platforms that continuously learn patterns,” he said.
This transition is also changing how cybersecurity is positioned within organizations. What was once seen as discretionary IT spending is now becoming embedded infrastructure, especially in highly regulated sectors. Majumdar said, “In sectors like BFSI… AI security is becoming as fundamental as core banking itself. Platforms like these will sit inline with transaction flows, not as overlays.”
The emergence of AI-on-AI cybersecurity is also reshaping how risk is assessed. With both attackers and defenders using intelligent systems, traditional underwriting models based on historical breach data are losing relevance. While platforms that enable real-time detection and response can reduce risk by cutting reaction times to milliseconds, they may also introduce new forms of systemic vulnerability if widely adopted systems fail.
Mishra emphasized that the effectiveness of autonomous defense systems will depend on three factors: speed, security and governance. “To make autonomous AI defense effective, three things are critical: first, speed… second, native AI security; and third, complete governance of the entire ecosystem,” he said, adding that even defensive AI systems must be protected against adversarial attacks.
Also Read | Explained: What is Mythos AI and why it is worrying banks and regulators
Boardroom conversations, particularly in BFSI and other critical infrastructure sectors, are already reflecting this shift. The focus is moving towards building resilient, self-defending systems capable of anticipating and neutralizing threats before they materialize.
As financial institutions double down on digital transformation, AI is no longer just a tool for fraud detection or threat intelligence—it is increasingly the battlefield itself. In this emerging paradigm, intelligent systems are not only defending infrastructure but continuously learning from equally intelligent adversaries, redefining the future of cybersecurity and investment in the sector.
