Charting AI's Path Forward In The Historical Context Of Disruptive Technologies That Came Before

*The views and opinions expressed by Anil Kumar are their own and do not necessarily represent those of any former or current employers.

Behind the promise of AI lies a surprising reality for banks. While most race to adopt it, their efforts remain largely fragmented. Now, disconnected pilot programs are creating more organizational friction than strategic advantage. The result is a significant challenge for banking leaders: turning scattered AI experiments into a unified strategy.

For an expert's perspective, we spoke with Anil Kumar, Senior Technology Lead at Barclays Corporate & Investment Bank. An ITIL-certified financial services IT professional with over 18 years of experience, including senior roles at Morgan Stanley, Kumar is deeply familiar with how large enterprises navigate technological disruption.

"The core challenge for leadership is aligning the fragmented efforts that define this inception stage. It's about moving beyond scattered pilots to build a unified strategy," Kumar said. But today’s chaotic environment is not unprecedented, he continued. Instead, it follows a familiar "crawl, walk, run" pattern seen in previous technology shifts.

• A familiar path: Rather than viewing AI as a unique crisis, financial institutions should see it as another chapter in a well-known story of technological disruption, Kumar said. "AI will follow a similar path to other disruptive technologies. We saw Bitcoin face initial skepticism before evolving with regulation. And just as most companies landed on hybrid models for the cloud, we'll likely see the same with AI."

Now, the industry is squarely in the “crawl” stage, he explained. One defining feature of this phase is the temptation to “bolt on” vendor tools to existing systems, a practice that weakens infrastructure and collects technical debt.

• The meaning of maturity: The goal is more than AI implementation, Kumar said. For him, success is measured by tangible business value, not technical sophistication. "Real maturity isn't about having the most advanced model. It’s about when AI consistently improves efficiency, reduces operating costs, and protects company data. That’s the goal, and we’re still about two to three years away from that being the norm."

The rush to adopt also introduces immediate dangers, Kumar cautioned. LLMs can be confidently wrong, for example, a risk that is especially pronounced in highly regulated sectors like finance.

• High-stakes risks: When combined with the fallibility of current models, an uncoordinated rollout can create significant consequences, he continued. "The security, compliance, and regulatory challenges are huge, and they won't be solved overnight. We're already seeing high-profile lawsuits over leaked content, like the one involving The New York Times. These models can replicate protected data and be misused in ways we're still discovering."

But how do leaders develop a strategy when the technology and risks are still being defined in real-time? For Kumar, success is about embracing ambiguity and adopting an agile mindset. "There is no simple formula right now. But you cannot let that uncertainty lead to inaction, because the risk of falling behind is too great. The only viable strategy is to experiment, measure the impact, and adjust your approach accordingly." To manage this uncertainty, he recommended using small successes to build momentum. Showcasing "quick wins" like a new chatbot or note-taking tool can create the organizational buy-in needed for the longer journey, he explained.

Ultimately, the competitive pressure forcing every organization to find its footing is the same force driving strategic fragmentation, Kumar concluded. "The market is definitely shifting toward AI. Every sector is being impacted. It’s become a kind of race."