BPI Leads Answer Reach While BDO Leads First Mentions in Philippine AI Banking Study

BPI appeared most widely across consumer, business and trust questions, while BDO was named first most often across all four tested question groups in the Philippine findings of the PRecious AI Authority Index: Southeast Asia Banking 2026.

Released by PRecious Communications on October 15, the research distinguishes between answer reach—whether a bank appears in an answer—and first-mention prominence—whether it is named before other banks. The findings describe the tested AI answers and do not rank banks by quality, safety, financial performance or customer preference.

Appearing in an answer and leading it are different measures

In Philippine business-banking questions, BPI appeared in 96% of tested answers, while BDO was mentioned first in 75% of responses.

The distinction also appeared in consumer banking: BPI recorded 73% answer reach, while BDO led first mentions at 31%. For trust-related questions, BPI’s reach was 92%, while BDO’s first-mention share was 64%.

The figures measure different aspects of visibility and should be read separately.

“The Philippine result makes the distinction very clear: visibility and prominence are different jobs,” said Paolo Alba, Regional Vice President and Philippines Country Lead at PRecious Communications.

“Communications teams need to know not only whether the bank appears, but what is said, why it is placed there and whether the supporting story reflects the experience customers actually have.”

Different AI routes can produce different answers

The regional study covered Indonesia, Malaysia, the Philippines, Singapore, Thailand and Vietnam, using 11 service-labelled test routes.

Across 1,628 eligible matched comparisons, the two routes answering the same question placed a different bank first in 706 cases, or 43%.

The study also found that business-banking answers named an average of 4.73 banks, compared with 3.52 banks in consumer-banking answers. These are regional findings rather than Philippine-only results.

For communications teams, the differences show why checking one service or one generic brand question provides only a partial view of how an institution appears in AI-generated responses.

What the research can—and cannot—show

PRecious describes the index as a diagnostic framework for examining brand descriptions, supporting public evidence and information gaps.

Most comparisons used application programming interfaces, or APIs, and source conditions varied. The study’s Copilot-labelled route used an Azure model with web search, rather than Microsoft 365 Copilot. Its Google route used a search-grounded Gemini API, rather than Google AI Overviews or AI Mode.

The results therefore describe those tested routes and should not be treated as universal behavior across the consumer products carrying similar names.

“First mention in an AI answer is not an endorsement or a measure of customer preference,” said Lars Voedisch, Founder and Group CEO of PRecious Communications.

“It indicates narrative order. Philippine communicators must examine what AI says about their institution, not just how often the name appears.”

Checking the evidence behind visibility

PRecious recommends starting with questions that matter to customers, testing relevant services and languages, and reviewing answers for inaccurate, incomplete or unsupported claims.

The research also distinguishes the roles of company-owned information and independent reporting. Both can contribute to the evidence surrounding a brand, but the study does not establish that media coverage causes inclusion in an AI answer.

The findings are a point-in-time snapshot. Changes to services, search results and public information can alter subsequent answers.

The report and methodology overview are available through the PRecious AI Authority Index website.