AI at a Crossroads: From Experiments to Economics

TL;DR – AI at a Crossroads

• Investments ≠ Returns: Many Filipino enterprises are still in pilot mode, exploring how to turn AI experiments into measurable business value

• Different Teams, Different Goals: IT focuses on efficiency, operations on workflow, but practical data strategies unlock value across departments.

• Escape PoC Purgatory: Use the DFV framework—Desirability, Feasibility, Viability—to prioritise projects that can scale.

• People First: AI works best when it enhances customer experience and empowers knowledge workers; efficiency follows naturally.

• Actionable Impact Wins: Start small, iterate fast, leverage imperfect data, and scale initiatives that deliver real business outcomes.

AI at a Crossroads: From Experiments to Economics

After years of investment, talent acquisition, and pilot projects, many Filipino enterprises are beginning to ask the harder question: where are the returns?

According to AIBP’s 2024 ASEAN Enterprise Innovation Report, 78.5% of ASEAN firms report moderate to significant AI investments as necessary to remain competitive. Yet many organisations remain at the “foundational” stage, dabbling in pilots, experimenting with copilots, and struggling to integrate AI into the messy realities of legacy systems.

To explore these challenges, AIBP, together with HCL Tech, brought together Filipino enterprises for a private workshop during the recent 50th AIBP Conference & Exhibition in Manila. Each organisation was at a different point in its AI journey, caught between enthusiasm and economics, seeking ways to move beyond proofs of concept and generate tangible business value.

The Many Faces of AI Ambition

Even within a single enterprise, different departments approach AI with very different expectations. During the roundtable, Leaders across industries are at a familiar reality: data is scattered, messy, and rarely ready for prime-time AI use.

One leader from a major Philippines Financial Institution shared that for their IT teams, AI often means wringing efficiency from repetitive tasks. 

“The simplest value“is always efficiency, getting more things done and letting repetitive tasks be taken care of.”, Eric Buenaflor, SVP Transaction Banking from Rizal Commercial Banking Corporation (RCBC). 

Where, the operations departments, on the other hand, might see AI as a tool for workflow automation.  As AI initiatives mature, approval processes have tightened: projects are no longer greenlit simply because they carry the “AI” label. As one participant put it: “I may not reduce costs, but I will deliver 20% more sales. That’s the game of business.”

Mustafa Sabuwala, Sales Director at HCL Technologies, suggests a pragmatic approach.

“Use the data in whatever format or framework. There are tools that will extract the data which will do pattern matching”, said Mustafa Sabuwala, Sales Director from HCL Technologies.

He explains that traditional approaches, forcing all data into a single master system, often collapse under their own weight. 

Rather than waiting for perfect systems, enterprises can start with the data they have, focusing on projects that deliver immediate value.

Escaping Proof of Concept Purgatory: Projects to Production

PoC Purgatory rears its ugly head when a company engages in a paid pilot or proof of concept that ultimately leads nowhere. Many enterprises face this challenge, especially when AI adoption varies widely across departments. 

Bernadette Salinas, Chief Data Officer from Philippine Long Distance Telephone (PLDT, the country’s largest telecommunications and digital services provider observed that in her organisation  some teams are highly progressive, experimenting with personal AI tools, while others are only beginning their journey and hope to advance gradually. The real challenge, she explains, is moving the teams at the starting line closer to the pace of the more advanced groups to close the gap so the organisation can scale AI effectively.

To navigate these challenges, Mustafa highlighted a simple framework of DFV - Desirability, Feasibility, and Viability.

• Desirability: Does the business truly want this project, and will it deliver meaningful impact?

• Feasibility: Can it actually be implemented, considering legacy systems, data accessibility, or greenfield environments?

• Viability: Can the project be sustained long-term and continue to provide value beyond the pilot?

Under the DFV lens, enterprises can cut through the hype and focus on initiatives that truly deliver results. This approach helps them prioritise projects with real impact, avoid wasted effort and keep stakeholders on board.”

Elevating Experiences, Not Just Operations

AI in the enterprise is often framed as a tool for efficiency: automating tasks, reducing costs, and streamlining operations. But the real differentiator is whether organisations deploy AI to serve people, not just processes.

At Bank of the Philippine Islands (BPI), the guiding principle is clear: customer obsession drives business value. AI is primarily applied to enhance customer experience, with the belief that improvements here will ultimately boost metrics such as NPS and revenue. The impact isn’t always immediate; AI investments are iterative and require careful evaluation over time.

Crucially, the bank doesn’t cling to sunk costs. “We start from the front using AI to improve the customer experience and then work our way back towards operational efficiencies and integrated data,”, said Virginia Braga, VP of Growth Marketing and User Experience from BPI. Pivoting is always an option if tools fail to deliver value. The BPI team was recognised at the 2025 AIBP Enterprise Innovation Awards in the Data and AI category for its AI-powered insights and analytics systems.

HCL Tech echoes this approach, emphasising AI as a way to empower knowledge workers. Tasks that once required teams of analysts, like research, report preparation, or data analysis, can now be handled by individuals using large language models (LLMs)

“I don’t think it will lead to the elimination of jobs but it is definitely leading to giving you back more personal time for you to think and act”, assures Sandeep Sarkar, Senior Vice President, Region Head ASEAN & North Asia from HCL Technologies. 

Turning AI Potential into Business Reality

The path from AI ambition to tangible value is rarely linear. It requires pragmatism, patience, and a focus on both human and business outcomes. Successful enterprises start small, iterate fast, and treat AI as an enabler of decisions, not just automation. 

Those that align AI with real business metrics, embrace imperfect data, and empower teams to experiment stand the best chance of moving from pilots to production and from promise to measurable results.

If thought-provoking discussions like this resonate with you, we invite you to explore our upcoming Philippines workshop in November. Over three days (November 18-20), we’ll cover the full journey from envisioning the future state of your organisation, to experimenting and aligning business processes, and finally to crafting the experience and managing change. Register your interest here.