MANILA, Philippines – NAICRI stands for the National Artificial Intelligence Center for Research and Innovation, a government program headed by the Department of Science and Technology (DOST) that seeks to “harmonize” the Philippines’ efforts to harness AI as a “force of development”. 

Officially launched on Thursday, February 26, with the DOST-ASTI (Advanced Science and Technology Institute) as the implementing agency, DOST Secretary Renato Solidum explained that while there have been many impactful project-based AI efforts before its establishment, these were fragmented and siloed. 

DOST is building NAICRI under a philosophy of a unified framework among the government, private firms, MSMEs (micro, small, and medium enterprises), and the general public with their various needs such as healthcare and disaster assistance. 

Solidum calls NAICRI as the “institutional anchor” to the National AI Strategy that was approved by President Ferdinand Marcos Jr. last year, May 2025. 

Financially, NAICRI takes part of an overall budget of P2.6 billion for its AI projects — described to be “relatively meager resources” but one that the department also hopes to scale by proving the effectiveness or benefits of its current and future programs. 

This amount may also be potentially augmented by potential grants by institutions such as the World Bank, said DOST-ASTI Director Franz de Leon, but stopped short of revealing specifics. 

What it seeks to address: AI skills, compute shortage

The ambitious program seeks to address “structural issues” namely a shortage of AI skills and talent; a need for more “compute” or computational capacity driven by data centers, and hardware including CPUs, and GPUs; and the need to solve “governance frameworks that have yet to catch up with the pace of technology” such as paper-based transactions. 

Solidum specified a need for a “26-fold increase” in nationwide compute power by 2028 as an “insufficient computing infrastructure risks the compute divide” — which may have the effect of the Philippines lagging in industries worldwide. 

De Leon told Rappler that aside from acquiring the necessary hardware from global partners or through partnering with local private firms, there’s also ambition shared with the local semiconductor industry and researchers in the academe — such as those at the University of the Philippines’ Electrical and Electronics Engineering Institute — to potentially design and build our own computer chips. 

Currently, the Philippines’ semiconductor industry is focused further down the value chain, specializing on chip assembly, testing, and packaging capabilities.

The DOST-ASTI’s approach to compute acquisition will be deliberate. “The last thing that we want to happen is that we buy certain servers or equipment. After 2 or 3 years, it will be outdated because new hardware will come up. Better, faster, more power efficient. If we have invested heavily at the front, it will be a waste,” De Leon said.

Through NAICRA, Solidum said, “We are building the shared computing backbone that allows Filipino researchers, agencies, and enterprises to train models, run simulations, and deploy AI applications at a scale that was previously out of reach. This means our researchers will no longer be limited to small-scale experiments.” 

“Through programs like GATES and ACABAI, we are developing a critical mass of Filipino experts, not only technical specialists but also domain scientists, ethicists, and public sector leaders who understand how to persuade AI into real-world impact,” he added. 

What are its big-ticket programs and current projects?

GATES and ACABAI are two more acronyms in a sea of them surrounding NAICRI — and two of three so-called pillars and “big-ticket programs” propping up NAICRI. 

GATES is Geospatial Analytics and Technology Solutions, and is a system being used to harmonize “fragmented data sets” to produce something more useful. For years, many agencies “generated valuable geospatial data, disaster, hazard, and risk maps from climate data, and agricultural data.” But they were siloed from one another. GATES is said to “harmonize all these datasets” to produce more impactful output. 

ACABAI is Advanced Computing, Analytics, Big Data, and Artificial Intelligence in the Philippines described as a program representing “the country’s most ambitious investment in AI…. [which] consolidates our AI platforms, use cases, and engineering capabilities under one umbrella.”

PROPEL, the third pillar, isn’t an acronym for anything. Launched in December 2024, the PROPEL program is the DOST’s “commercialization arm” that is focused on making sure that innovations and output under GATES, ACABAI, and other AI initiatives actually become usable, beneficial or market-ready for the general public. “Make sure research goes beyond the laboratory, and actually reaches Filipinos,” Solidum said. 

The DOST showcased a few other already operational innovations at the launch such as:

1. vBantai (Vision-Based Adapative Intelligent Traffic Control with Artificial Intelligence) is a vehicle-recognition model able to identify local vehicles accurately such as jeeps and tricycles. Internationally-made systems previously misidentified, for instance, jeepneys as trucks, making the data not as useful for traffic analysis. This will be used by agencies such as the Metropolitan Manila Development Authority (MMDA) for traffic analysis, currently done manually, with the hope of easing the massive traffic problem.
   
2. Mus3o – AI-enabled digitization of cultural and natural heritage assets such as animal specimens and historical artifacts using computer vision, imaging, and 3D reconstruction workflows. 
   
3. AI4RP (AI-Powered Weather Forecasting for a Resilient Philippines) – Also known as Project Gabay, the AI-enabled system allows forecasts at “10 times higher resolution” that cuts forecast generation to 15 minutes from 3 hours before. It is trained on local meteorological data, and was deployed in 2024.
   
4. ROAMER (Robot for Optimized and Autonomous Mission-Enhancement Responses) – A prototype autonomous unmanned ground vehicle designed to track, survey, and map banana farms, which is said to “improve path planning for chemical administrations such as fertilizers.” 

The NAICRI program’s immediate next steps include holding a national survey to develop an “AI compute profile” to determine capacity needs not just in the capital but also in the regions, the development of an AI training catalogue, and training for MSMEs and regional institutions on how to tap into the DOST’s computing facility and backbone. 

Erika Legara, chief AI and data officer at the Center for AI Research, added context as to AI’s function: “We are not building AI for the sake of building AI. AI is really a tool, so it’s important we identify first where it can help best. We won’t just create just to be able to say that we have AI tools.” 

The World Privacy Forum lists 95 countries and jurisdictions with a national AI strategy plan, and 139 without. – Rappler.com