Entersekt CIO: Banks Need AI as Antiviral, Not Feature, to Combat Fraud

Banks have debated whether artificial intelligence represents an enhancement or a competitive edge, but Entersekt's chief information officer Richard Bailey argues the conversation starts elsewhere. AI has moved beyond the traditional aspirin-or-vitamin framework into a third category: an antiviral that continuously adapts as fraud evolves.

Financial institutions approach Entersekt seeking relief from immediate challenges, not aspirational innovation. Bailey emphasized that his customers face acute problems around disputed payments, secure onboarding, and attacks that shift faster than defenses can adjust.

Buyers now expect measurable improvements in authentication, fraud mitigation, and customer experience rather than technology for its own sake. They evaluate solutions based on whether AI demonstrably moves key performance indicators, not whether it simply sits atop an existing platform.

Attackers have gained access to identical technologies, fundamentally altering the threat landscape. Bailey noted that fraudsters can now enrich their attacks and deploy multiple vectors simultaneously using AI.

Static rules and periodic updates struggle against adversaries capable of generating convincing synthetic identities or deploying deepfakes at scale. Voice and facial biometrics lose effectiveness when AI enables convincing impersonation.

If AI is using deepfakes, for example, then voice biometrics and facial biometrics doesn't work anymore. We might need to look for another way of defending against that type of attack.

The antiviral analogy reflects this reality. Rather than treating symptoms or boosting general resilience, AI must constantly recognize emerging threats and adjust defenses during live transactions.

Bailey challenged the notion that copilots or chatbots constitute meaningful transformation. The critical shift involves whether AI functions as a product layer or becomes part of the underlying infrastructure.

A support assistant that summarizes fraud cases may improve efficiency, yet infrastructure-level AI influences authentication decisions, recommends policy changes, and responds to new attack patterns before losses mount. The industry approaches a threshold where AI becomes integral to product infrastructure, enabling organizations to trust it with operational decisions inside regulated environments.

The real shift is not really from aspirin to vitamin or vitamin to aspirin. It's from is AI a feature of the product layered on top of the product … or is it now part of the infrastructure of the product?

Disconnected products assembled from multiple vendors may prove insufficient because attractive interfaces can be replicated quickly. Durable advantages emerge from capabilities that are considerably harder to reproduce.

Financial institutions must explain AI decisions to regulators and auditors while certifying compliance with industry standards. Governed information creates a second advantage — relevant, clean, properly managed customer data provides context that generic AI systems cannot easily duplicate.

Bailey argued that models may be replicable, but the data underpinning them is not. His closing observation tied these themes together: the pivotal decision for buyers is no longer whether they can use AI, but whether they can use AI to run the business.