When data becomes contested, trust must be built

We are living through a strange contradiction. Never before have societies produced so much data, and never has it been so difficult for the public to know what information deserves trust.

Data is being collected at extraordinary speed. Governments, companies, universities, journalists and civic organizations use it to explain economies, elections, public health, climate, education and daily life. Yet the same information can be selectively released, stripped of context, manipulated for influence or hidden behind systems that most people cannot inspect. In the worst cases, data is altered, erased or weaponized. In quieter cases, it is simply made inaccessible.

This is what we mean when we say data is under attack.

The attack is not always dramatic. Sometimes it appears as a misleading graph shared thousands of times. Sometimes it is a model whose assumptions remain undisclosed. Sometimes it is a public database that becomes harder to access, a methodology that changes without explanation or a headline that presents uncertainty as certainty. Each instance may seem small, but together they weaken the public’s ability to understand reality and make informed choices.

That is why we built Arkipelago Analytics.

We created Arkipelago Analytics to help protect data, strengthen its credibility and deliver it to the public in a form that is clear, useful and accountable. We believe reliable information should not belong only to specialists or institutions with the resources to interpret it. Data has public value, and that value increases when people can see where it came from, how it was analyzed and what its limitations are.

Our commitment begins with transparency.

Transparency means showing our work. It means identifying sources, documenting methods and explaining the assumptions behind an analysis. It means distinguishing between what the evidence demonstrates, what it merely suggests and what remains unknown. It also means correcting the record when better information becomes available.

This matters because data science is not an automatic truth machine. Every analysis contains human decisions: which questions to ask, which variables to include, which populations to count, which comparisons to make and which outcomes to emphasize. Credible data science does not hide those decisions. It makes them visible and open to scrutiny.

Our commitment also extends to accessibility. A technically sound report can still fail the public if it is written in language few people understand or presented in a format they cannot use. We want to translate complex evidence without distorting it. That requires careful communication, thoughtful visualization and respect for the audience. Simplicity should clarify the truth, not flatten it.

The name Arkipelago reflects how we see the information landscape. Valuable data often exists in separate islands: scattered across agencies, disciplines, communities and platforms. Our work is to connect those islands while preserving their context. We bring sources together, test their reliability and help people see the larger picture without pretending that every gap has been filled.

Trust cannot be demanded. It must be earned through repeated acts of openness, rigor and honesty. We do not expect the public to accept a conclusion because it is presented by an analyst or displayed on a polished dashboard. We want people to understand the evidence well enough to question it, use it and hold us accountable for it.

More data will not automatically create a better-informed society. That will require institutions willing to defend integrity over convenience and clarity over persuasion.

Arkipelago Analytics was built for that purpose: to protect data, practice credible data science and ensure that trustworthy information reaches the public. When facts are contested and confidence is fragile, transparency is not simply a principle. It is a public service.

-Atty. Anna Mae Yu Lamentillo is the Chief Technology Officer of Paraluman News  and the founder of Arkipelago Analytics, Build Initiative Foundation, and Night Owl AI.