Why the Philippines must prepare for AI-driven data breaches

FILIPINOS love the internet. With 97.5 million people online, equivalent to 83.8% of the population, the Philippines ranks among the heaviest internet users in the world. Nearly all of these internet users (98.5%)

- By Ram Vaidyanathan

access it through their mobile phones, spending close to nine hours online daily. About five of those hours are spent on mobile devices, with the remainder spent on desktops and tablets.

From social media posts and photos to professional updates, the data trail left behind is enormous. What many do not realize is that these seemingly harmless digital footprints, such as holiday albums, casual social media posts, status updates, or even job profiles, are increasingly being tapped to train artificial intelligence (AI) systems that now power businesses and applications worldwide.

And here lies the problem. Once data enters an AI model, users lose control over how it is used, meaning it can be replicated or even misused.

THE INTERNET NEVER FORGETS

The internet has a long memory, and everything is archived online.

All of this accumulated data, stretching back years, can now be used to train powerful machine-learning models. Disturbingly, the scope of what is being ingested is not always transparent.

Take a recent case uncovered by Proof News, a nonprofit newsroom. Its investigation revealed that some of the world’s largest AI companies had quietly harvested data from over 173,000 YouTube videos across 48,000 channels, despite explicit rules prohibiting this kind of extraction. Tech giants such as Anthropic, Nvidia, Apple, and Salesforce were all named in the report.

The volume of data that companies gather from us, and how they use it, is often buried deep in the fine print. Most people click accept without realizing just how much personal or behavioral information they are handing over.

LOCAL RULES, GLOBAL STAKES