Hoy, kumusta? Welcome back to Kaisipan!

Pax Silica is no longer just a declaration. Specific companies are named, deals are moving, and the US is already in Manila. The question now is who actually wins.

We've been tracking Pax Silica since the US first pitched it to the Philippines in December 2025. The Philippines joined officially in April 2026. Now, the deals are getting specific, and so is the fine print.

The diplomacy phase is over. US Undersecretary Jacob Helberg arrived in Manila this week to push implementation of the Pax Silica commitments, including a visit to the proposed Economic Security Zone in Luzon.

DTI Undersecretary Rodolfo wrapped a high-level tour of the UAE, Israel, and Türkiye to firm up partnerships for the 4,000-acre Luzon AI hub. Two UAE deals are now concrete:

Side meetings in Israel and Türkiye covered critical minerals processing and trade revival: supporting pieces, not the headline.

Here's the government's pitch. Electronics and semiconductors make up roughly 60% of all Philippine exports, worth about $36 billion a year, per SEIPI. But the country's role has always been confined to low-value work: assembly, testing, packaging. Pax Silica is meant to push the Philippines up toward chip design and advanced manufacturing. At an April 24 press conference, BCDA CEO Joshua Bingcang said membership would generate over 100,000 jobs.

But some of the fine print has now surfaced, and it deserves scrutiny. According to Rappler, the Economic Security Zone will operate under US common law, not Philippine law. It carries diplomatic immunity similar to a US embassy. The Philippines offers the land rent-free for the first two years, after which the arrangement can be renewed for up to 99 years.

Opposition from the Makabayan bloc and KMP (Kilusang Magbubukid ng Pilipinas), a national farmers' movement, was covered in our last issue and continues. The fresh voice this week: Kalikasan PNE (People's Network for the Environment), an environmental advocacy group, warned that the mountains, watersheds, and ancestral lands supplying the hub's raw materials face irreversible ecological harm. They've also called for congressional scrutiny and full public disclosure of the terms.

A deal where Philippine land operates under US law, with 99-year renewal terms, is not just an economic question. It's a sovereignty one.

According to KnowBe4's Philippine threat report, phishing attacks in the Philippines jumped 423% between 2024 and 2025. The driver: AI-powered, industrialized scams targeting a mobile-first population. Criminals shifted to high-volume smishing (SMS-based phishing) to bypass traditional security filters, and used AI to mimic legitimate brands, institutions, and even individuals.

The broader damage is already staggering. The Philippines is the second most-breached country in Southeast Asia, with over 151 million user accounts compromised since 2004. Social engineering, tricking people into handing over credentials or money, accounted for more than three-quarters of all financial fraud in 2025.

What makes this harder now is AI's ability to personalize attacks at scale. Threat actors analyze publicly available data to craft messages that feel personal and authoritative. The old tells are gone: bad grammar, misspelled logos, generic greetings. These attacks now read like they came from your actual boss.

BPO workers, freelancers, and remote employees are especially exposed. One convincing fake email from "HR" or "the client" can compromise an entire system. Malicious email attacks rose 15% globally in 2025, with Asia-Pacific recording the highest share of email antivirus detections at 30%.

The Philippines is expanding its digital footprint rapidly, as the Pax Silica deals clearly show. But going digital without a serious security culture is an open invitation. AI is raising both sides of this equation at once: making innovation faster and making exploitation easier.

80%. That's how many Philippine CEOs said they are comfortable making major strategic decisions based on AI-generated insights, according to the IBM Institute for Business Value, which surveyed 2,000 CEOs globally. The global average: 64%. Philippine executives are among the world's most confident in using AI for high-stakes calls.

The numbers go further. 73% of Philippine executives believe AI makes tactical and operational decisions faster and better than humans, compared to 57% globally. By 2030, Philippine CEOs expect AI to handle nearly half of all operational decisions without human oversight, up from 25% today.

That's not just confidence. That's a fundamental restructuring of how leadership works. Globally, the share of organizations with a Chief AI Officer jumped from 26% to 76% in a single year, per IBM's full report. Philippine organizations are moving the same direction, with Chief AI Officer influence expected to expand further through 2030. About 53% of Philippine executives also expect HR leaders to gain more influence precisely because of the workforce disruption AI brings.

97% of Philippine CEOs in the study also said business leaders must become technology experts within their own functions. Hindi na lang basta "tech guy" ang gagawa ng AI strategy. Lahat kailangang marunong.

But here's what deserves a second look: 73% of Philippine executives admitted they are investing in AI out of fear of falling behind competitors, even without fully understanding the business value. That's not strategy. That's FOMO with a budget.

And the workforce consequences are arriving fast. Between 2026 and 2028, Philippine companies expect 31% of employees to need reskilling for entirely new roles, while 55% will need additional training just to keep up with their current positions. Around 90% of Philippine CEOs say AI success ultimately depends more on employee adoption than on the technology itself.

Personally, I find this confidence gap worth watching. Gawing AI aid ang decision-making? That's the right approach. Pero kapag sinabi ng CEO na mas matalino pa ang AI kaysa sa kanya, at aminado namang hindi niya lubos naiintindihan ang value nito, that's not strategy. That's delegation to a black box. AI is a tool. It helps you think better. It should never be the one doing the thinking.

The portrait of Filipino business leadership on AI is this: bold, optimistic, and self-aware enough to see the gap between ambition and execution. Whether companies close that gap before their employees fall behind is the real test.

54 seconds. That's how fast a researcher at the University of Regensburg wrote up a fake scientific experiment, one he never actually performed, using an AI tool from OpenAI. His post on Bluesky doubled as a warning: writing a paper has never been easier, and neither has clogging the research pipeline.

Nature magazine flagged this as a growing crisis in academic publishing. Preprint repositories, platforms where researchers share work before formal peer review, are being flooded with AI-generated submissions: papers with fabricated experiments, made-up citations, and research that was never actually done. ArXiv, one of the world's largest scientific preprint repositories, recently announced stricter policies to push back against the wave.

Why should Filipinos care? Because research shapes decisions. Medical protocols, educational policies, government programs: these rely on academic literature. If that literature is being contaminated by AI-generated content designed to look credible without being truthful, the downstream effects reach far beyond journal editors and conference chairs.

The Philippines has a growing research sector. Universities, think tanks, and government agencies increasingly turn to international literature to inform local decisions. Knowing that this literature is now under threat from "AI slop," meaning low-quality, AI-generated content that mimics rigorous research, is relevant information.

AI makes producing coherent-looking text effortless. For science, where truth matters more than coherence, that's a structural threat.

Honestly, arXiv's crackdown makes complete sense. Research is one of the few spaces where AI involvement should be minimal or strictly limited. Using AI to handle complex mathematical calculations or run simulations? That's a valid tool. Pero kung ang buong paper, pati na ang argumentation at conclusions, ay AI ang gumawa — I won't trust that research. And neither should you. The value of a study lies in human judgment applied to real evidence. AI can assist that process. It cannot substitute for it.

A prankster posted a real Monet painting on Twitter and told everyone it was AI-generated. The replies were spectacular.

"No coherence of elements." "Emotionless." "It's garbage." Confidently wrong, every single one of them. No actual analysis. No actual looking. Just a posture. The assumption was simple: AI-generated, therefore bad. That was enough to open the floodgates.

Then the reveal came. A real Monet. Many deleted their posts. Napahiya.

And here's what bothers me most. Of all the people who piled on, did even one of them actually know Monet's work? A real art historian, a genuine expert in his style, would have caught it. The brushwork, the light, the water lilies — these things have tells that no amount of AI can fake at the detail level. A true Monet expert would zoom into the smallest corner of that canvas and know. But the people who commented? They weren't experts. They were rage-baiting. Farming likes and follows by performing outrage at AI. That's all it was.

And this is the real lesson. Expertise is king in the age of AI. Not hot takes. Not performative anti-AI sentiment dressed up as taste. Actual, deep knowledge of a subject is the only thing that holds up when the lines start blurring. Because when you don't have it, you end up confidently wrong on the internet — and then quietly deleting your posts when the truth comes out.

If you hate AI, hate it. That's your right. Own it as a principle, not a performance. But projecting insecurities onto others while pretending to have taste? That's the move of someone who clearly couldn't tell a Monet from a chatbot.

Thank you for reading today’s newsletter!

Do you have suggestions? I’d love to hear your thoughts.