USING ARTIFICIAL INTELLIGENCE TO TRACK CORRUPTION IN PUBLIC WORKS

USING ARTIFICIAL INTELLIGENCE TO TRACK CORRUPTION IN PUBLIC WORKS

It is interesting to note that President Ferdinand R. Marcos, Jr. has recently pointed out a peculiar pattern in the data—most likely from the Department of Public Works and Highways (DPWH)—showing that many flood control project contracts had the same project costs. Almost like they came out of a cookie cutter.

Equally curious, the President discovered that only about 15 contractors have been bagging most of these projects. Now, this is not just a random coincidence. This is a red flag that, if ignored, could allow corruption to keep flowing as freely as floodwaters in an unprotected barangay.

I am encouraged that such data sets are reaching the President’s desk and that he is reading them—and more importantly—sharing them with the public. That is already a big first step. But here’s my unsolicited advice: the next step should be to use artificial intelligence (AI) and machine learning (ML) to dig deeper into this data. And while we’re at it, let’s widen the net to catch even more patterns that the naked eye can’t detect.

Where should he get more data? Aside from DPWH, there’s the Department of Budget and Management (DBM) and the Commission on Audit (COA). Imagine this: data from all three agencies, juxtaposed, cross-referenced, and run through AI-powered analysis. I’m willing to bet my morning coffee that the patterns will start jumping out like fish in shallow water.

And why stop there? Let’s add the databases from PhilGEPS (Philippine Government Electronic Procurement System) and the Anti-Red Tape Authority (ARTA). In fact, if we feed AI the Terms of Reference (TORs) for all projects, we might uncover “custom-tailored” specifications designed to favor a chosen contractor.

This is not science fiction. Governments around the world are already using AI to spot procurement anomalies, detect price rigging, and even identify shell companies hiding behind multiple layers of ownership. AI can:

1.   Monitor procurement – Scan bid data to find repeated winners, suspicious pricing, and unusual bidding behavior.

2.   Flag high-risk transactions – Assign risk scores to contracts and vendors, based on past patterns.

3.   Map hidden networks – Reveal connections between people, companies, and even offshore entities.

4.   Assist audits – Help COA or internal auditors comb through massive records faster.

5.   Enable real-time alerts – Catch irregularities as they happen, not years later in an audit report.

This is especially powerful when combined with tools like Benford’s Law, which detects unnatural number patterns in financial data—an old auditor’s trick that still works wonders when scaled up with AI.

Of course, all this requires clean and interoperable government data. It also demands proper legal safeguards to protect privacy and ensure due process. AI must assist human investigators, not replace them. No algorithm should have the final say in accusing someone of corruption—but it can point us toward where to look.

Now here’s where the Philippines can take a leadership role. If Indonesia and Malaysia can already partner with international innovators like The Ocean Cleanup for environmental work, why can’t we form similar partnerships for governance technology? We can invite global AI anti-corruption experts to collaborate with our agencies, train our analysts, and set up pilot programs starting with high-risk sectors like public works.

I could immediately mobilize a “ragtag army” of volunteer ICT experts that would gladly help the President in this mission. We have the talent. We have the data. And we certainly have the motive—because corruption doesn’t just waste money; it kills projects before they can save lives.

If we can use AI to predict typhoons and track pandemics, we can use it to track corruption. Imagine an app, accessible even to barangay officials, where suspicious transactions are flagged in real-time. Imagine the deterrent effect when contractors know that every peso they touch is being watched by both human eyes and machine learning models.

Corruption thrives in darkness. AI can be the floodlight.

Ramon Ike V. Seneres, www.facebook.com/ike.seneres
iseneres@yahoo.com, 09088877282, senseneres.blogspot.com

10-20-2025