The more AI banks deploy, the bigger the cyber battlefield gets - The Economic Times

TEXT START: This comes as the financial sector itself races to deploy AI across lending, payments, fraud detection and customer service, creating an emerging cybersecurity arms race: as attackers use AI to probe systems faster, financial institutions are turning to AI to detect, predict and respond to threats at a similar speed.

The Dissection

The article accurately maps AI’s expansion of the cyber battlefield: automated attacks, autonomous agents, machine identities, excessive permissions, shadow AI and legacy-system exposure.

But its deeper function is normalization. It treats AI deployment as inevitable and presents the consequences as an engineering and budget problem: spend more, govern access, deploy defensive AI, improve literacy. The structural rupture is reduced to an enterprise risk-management checklist.

The article’s most revealing admission is that if the attacker is a model, the defender cannot remain human. That is a direct concession to Cognitive Automation Dominance. Modern finance is being pushed toward machine-versus-machine operation because human cognition cannot match the speed, scale or persistence of the systems it has created.

The Core Fallacy

The article mistakes controllability for reversibility.

Using AI to defend against AI may reduce specific vulnerabilities, but it does not restore the post-WWII economic circuit. It does not preserve mass employment, human productive necessity or broad access to ownership. It merely hardens the institutions that are automating those functions.

The “AI for AI” solution also assumes a manageable symmetry between attack and defense. There is none. Attackers need one exploitable weakness; defenders must contain every relevant weakness across thousands of entry points, identities and dependencies. AI accelerates both sides, but the asymmetry remains—and the cheapest, fastest and least constrained actors gain disproportionate leverage.

Under the Discontinuity Thesis, the result is not secure continuity. It is more concentrated control over financial infrastructure by whoever owns the strongest models, data, compute and defensive systems.

Hidden Assumptions

• Banks will remain stable, legitimate institutions while AI removes the labor that historically connected wages to consumption.
• Cybersecurity is primarily a spending deficit rather than a structural consequence of autonomous systems with expanding authority.
• Human executives, regulators and security teams can meaningfully understand and control agents operating across legacy systems, cloud platforms and third-party identities.
• A common governance layer can contain complexity without becoming another concentrated point of failure or control.
• Defensive AI can keep pace indefinitely with offensive AI rather than becoming part of an escalating arms race.
• Permissions can be inventoried, constrained and revoked reliably even as machine identities multiply faster than institutions can audit them.
• AI agents will remain subordinate tools instead of becoming operational actors whose access determines who effectively controls the bank.
• Customer education can materially offset industrial-scale social engineering once attackers can generate and adapt deception continuously.
• Protecting banks is equivalent to preserving the existing economic order. It is not. A bank can survive as an automated fortress while its workforce becomes economically disposable.

Social Function

Classification: partial truth packaged as transition management and ideological anesthetic.

The article is not false. It identifies a real and worsening attack surface. Its anesthetic effect comes from framing the crisis as solvable through better spending, identity governance and faster deployment. That script allows institutions to claim adaptation while avoiding the more terminal question: who owns the automated financial infrastructure, and what happens to everyone no longer required to operate it?

It also performs elite self-exoneration. The institutions accelerating automation are portrayed as responsible actors forced to defend themselves against attackers, rather than as participants in a competitive process that makes human labor less necessary and systemic control more concentrated.

The Verdict

Accurate on the battlefield, evasive about the war.

The article documents P1 in action: cognitive work, including attack detection and defense, is being transferred from humans to models. It also points toward P2: no institution can preserve a stable human-only security domain once adversaries operate continuously at machine speed. The article does not directly establish P3, but its own remedy—machine defenders against machine attackers—shows where the system is going.

AI cybersecurity may preserve banks. It will not preserve post-WWII capitalism. It converts financial institutions into increasingly autonomous, concentrated control systems while the human workforce loses productive necessity. The battlefield gets larger because the economic order is building weapons and defenses it can no longer operate through ordinary human participation.