Philippine Companies Show Why GenAI Success Isn’t About the Models

It’s easy to get caught up in the hype surrounding AI, especially with the news about bigger models, more impressive benchmarks, and the latest seven-figure funding rounds. But what if the key to unlocking AI’s true potential isn’t about the models at all?

This was the conversation that occurred inside a Manila hotel boardroom, where senior executives from some of the Philippines’ largest conglomerates — spanning banking, telecommunications, retail, and real estate — became candid about the reality of generative AI.

Sponsored by AWS and held under the Chatham House Rule, the lively discussion held at the end of the half-day Chief Digital and Data Officer Philippine Summit dwelt on a major truth: The companies winning at AI are the ones who’ve cracked the code on three critical, non-technical challenges, and not necessarily the ones with biggest budgets or the most advanced technology. These challenges include license governance, organizational design, and navigating the human element of fear. By focusing on these three, they’re also solving the problem that’s stumping Fortune 500 companies worldwide: how to make AI actually work in the real world.

Owning the licensing problem

An immediate hurdle for any large company driving an AI agenda within is managing the flood of requests for AI tools. “For large companies, people are going to ask for AI, right? So how do you govern in such a way that not everybody needs it, but when everybody’s going to request it?” a CTO from the real estate development sector explained.

Licensing tools like Microsoft Copilot, which leverages large language models and connects with Microsoft 365 data and Microsoft Graph, can quickly become a seven-figure decision. However, this is where Philippine enterprises get creative. A multinational consumer goods company had a surgical approach to license management: They audit usage and reallocate licenses from non-users to power users every three months.

“What we’re doing is we have our regular audit using the tracker inside the tool, so you would know how often people use it,” a senior analytics leader says. “For those who are not using it for three months, we decide to take their license and give it to someone who really needs it.”

Another major Philippine bank took a more radical approach, making AI a budget line item for department heads. “We actually charge it to the different departments. And now we actually have business heads who are at least more critical to deciding whether they're going to give their people the license, because it’s now part of their cost," the senior IT executive from the banking industry explains.

These aren’t penny-pinching tactics. Instead, it is a fundamental rethinking of how AI tools are deployed. By doing so, Philippine companies are strategically throttling adoption to ensure it’s both efficient and impactful without blowing their budgets.

Getting honest with data-first strategies

While most AI vendors push a use-case-driven approach, some Philippine companies are doing something different: they’re starting with their data. “Check your data, and then build use cases, rather than try to adopt a lot of fancy use cases,” advises a senior technology executive from the infrastructure sector. “Sometimes [POCs] become part of an IT graveyard because there isn’t any understanding of what the data can do.”

This data-first strategy reveals other necessary truths. At an investment management firm, a technology leader discovered that critical portfolio management data was scattered across relationship managers’ laptops. “Data is everywhere. Now we’ve started to unify our own data warehouse, so all the data is now at a central repository.”

The key takeaway? Before you can even think about prompt engineering, you need to master data engineering.

AI success needs a human approach

The most crucial insight from the discussion is about tackling human emotion, i.e., fear. A technology executive from the property development sector captured a feeling that every AI implementation team knows but won’t admit: “Some of the employees are too cautious, because they will think AI might replace them. And that’s because they see some employees already being replaced by this AI.”

Successful companies aren’t ignoring this fear but are architecting around it. This same executive’s promise to his team was simple: “Nobody gets hurt, nobody will be removed.” This isn’t naivete but a deliberate strategy. When fear is removed from the equation, adoption accelerates, says the executive. “When nobody in the picture is going to be removed, or all of the people in the organization will stay, everybody cooperates.”

The discussion also addressed what economists call the AI productivity paradox — where AI tools create measurable productivity gains that don't always translate to bottom-line improvements. The solution isn’t more AI, according to the attending Philippine executives. It’s really about making the hard organizational decisions about workforce planning and where to grow the organization, like reskilling employees, restructuring departments, or creating new roles.

The future of enterprise AI: The Philippine edition

While the tech world obsesses over the next big AI thing, Philippine enterprises are quietly building an entirely different future. A senior director from a major telecommunications company revealed they were implementing Model Context Protocol (MCP) servers connected to their monitoring systems, allowing IT operations to use a chatbot to check if systems have reached certain thresholds. It also allows them to be more model-agnostic.

At the same time, Philippine enterprises understand that an AI journey cannot be a solo one. A CEO from the business process outsourcing sector found an unconventional partner: the AWS Innovation Center, which is now helping his team build a custom solution to analyze hundreds of video interviews after they were accepted into the program. It’s not about looking for infrastructure-as-a-Service anymore; it’s about becoming innovation-as-a-Service and working with AI partners who understand this.

The panel discussion showed that Philippine companies are not chasing artificial general intelligence (AGI) but are more focused on operationalizing the AI we already have. They see the future of enterprise AI as not technical but more tactical. And, the Filipinos are showing everyone else how it’s done.

Image credit: iStockphoto/-izabell-