OUR country is moving toward an unprecedented trifecta of elections. The first regular Bangsamoro Autonomous Region in Muslim Mindanao (BARMM) parliamentary elections are set for September 2026. The Barangay and Sangguniang Kabataan elections (BSKE) will follow in November 2026. Finally, the national and local elections (NLE) arrive in May 2028. Each event serves as a critical junction in our democratic system.
Yet, as we build the physical and digital infrastructure for these polls, a parallel infrastructure is being laid down by information manipulators. Our political system is reeling from a multibillion-peso flood control scandal, the ongoing impeachment process of the vice president, and deep polarization within the Senate. These are not mere headlines. To a systems analyst, these crises represent vulnerable entry points — exploitable system bugs — that domestic and foreign actors can weaponize to manipulate public perception.
With artificial intelligence (AI) becoming more sophisticated and accessible, the threat has evolved from simple fake news to automated, scalable influence operations. Understanding how these narratives are engineered is the first step toward building our national defense.
At the base of our administrative system lies the barangay. The BSKE is highly personal, fought community by community. In the past, running a disinformation campaign across thousands of unique villages required immense human resources. Today, generative AI solves the problem of scale.
The recent exposure of massive corruption in flood mitigation budgets provides the perfect raw material. Manipulators can feed local flooding data, municipal project registries, and specific regional dialects into AI models. The output? Thousands of unique, hyper-localized social media posts tailored for specific neighborhoods.
A voter in a flooded low-lying area might see a seemingly organic post written in their exact local slang, complete with AI-enhanced video, accusing their specific barangay captain of pocketing localized drainage funds. By automating the production of localized anger, bad actors can disrupt the BSKE and the BARMM elections without ever deploying a traditional, centralized troll farm. Because these posts use local contexts and dialects, they easily bypass standard content moderation filters used by major tech platforms.
As the political warfare in Manila intensifies around the impeachment trial, the sheer volume of conflicting reports creates severe data noise. This noise allows for the deployment of a dangerous concept known as the “liar’s dividend.”
When AI can create highly realistic deepfake audio and video, objective truth becomes a casualty. If a deepfake audio file of a senator negotiating political favors surfaces, it spreads instantly. By the time technical experts verify that the file is synthetic, the political damage is already done. Cognitive anchoring has occurred; the public has already processed the lie as truth.
Worse, the reverse is equally true. When genuine evidence of corruption or political maneuvering comes to light — such as financial documents or recorded admissions related to the infrastructure scams — the accused can simply claim that the evidence is an AI-generated fabrication. When everything can be faked, nothing can be trusted. This completely paralyzes public trust, turning the upcoming Senate trial into a theater of disbelief.
While domestic politicians use these narratives for immediate survival or local gain, foreign state-sponsored actors operate on a completely different layer of the system. Their objective is not to help a specific local candidate win, but to degrade public trust in our democratic institutions ahead of the 2028 NLE.
Foreign adversaries can exploit our domestic crises to craft a macro-narrative: that the Philippine democratic system is fundamentally broken, inherently corrupt, and structurally unviable. Using AI-driven botnets that do not just blast text but actively simulate human debate in online comment sections, they systematically amplify the most extreme viewpoints from both sides of our political divides.
The strategic goal is to induce civic exhaustion. If the public becomes convinced that every public official is corrupt, that the Senate is a playground for elite feuds, and that the automated election system cannot be trusted, public morale collapses. A nation turned inward by perpetual domestic distrust becomes fragile, distracted, and incapable of executing coherent foreign policy or mounting a unified territorial defense.
In the context of the BARMM elections, this foreign intervention can be catastrophic. By distributing AI-synthesized audio leaks or fabricated official memos within secure messaging apps, foreign actors can make regional factions believe they are about to be betrayed by national security forces or rival groups. In a post-conflict transition, information gaps breed intense paranoia. It takes very little synthetic evidence to trigger real-world security flashpoints on the ground.
How do we defend against automated deception? The traditional approach of manual fact-checking or attempting to police content is no longer viable. Human fact-checkers cannot scale at the speed of an algorithm that generates text and video at zero marginal cost.
Our national defense must shift from monitoring content to inspecting conduct and ensuring technical provenance.
First, we must institutionalize forensic integrity protocols. Every official government announcement, Senate transcript, and Comelec advisory must use cryptographic watermarking. This allows the public to instantly verify the origin of a digital file.
Second, our cybersecurity watchdogs must monitor behavioral markers rather than specific words. We must track automated account creation patterns, unnatural posting frequencies, and coordinated cross-platform timing anomalies to shut down botnets before they achieve virality.
Elections are the core protocol of our democracy. If we allow the data feeding that protocol to be systematically corrupted by sophisticated operations, the integrity of the entire system fails. Protecting our upcoming elections requires us to see through the political noise and secure our information infrastructure with technical rigor and absolute transparency.
