.jpg)
It is not enough to just "use" AI. To secure our future, we must transition from being passive consumers to sovereign architects of intelligence.
We start with what we have: deeptech expertise from the semiconductor industry, local philosophical frameworks (Adab & Governance), and a network of high-agency builders.
"If we combine technical AI safety knowledge with local cultural wisdom, we can create governance models that actually stick in the ASEAN context."
We deploy our "3-Prong Approach" to target the critical nodes of the ecosystem. We don't just talk; we build prototypes and run drills.
Deploying offline SLMs & tracking GPUs.
Cetalabs Kinder & Family Workshops.
Digitizing Sejarah for local datasets.
"By engaging students in 'Red Teaming' and families in 'Digital Guardrails', we demystify AI risks and turn anxiety into agency."
What happens after we run these programs? The ecosystem begins to shift from passive consumption to active management.
"Success looks like a student who knows how to 'jailbreak' an LLM to test it, and a policymaker who knows why we need a 'kill switch' for data centers."
The long-term vision. A Malaysia that is not a digital colony, but a safe harbor for aligned, human-centric intelligence.
"We build the brakes so we can drive fast, safely. We build the engine so we decide the destination."
How transformation ripples outward.
The individual moves from being a "user" to a "builder". They understand the tools they use, can detect manipulation, and have the technical skills to deploy local AI solutions.
Families act as the first line of defense against digital addiction. Communities form "trust networks" to verify information, protecting the social fabric from algorithmic polarization.
The nation secures its digital borders. We own our compute, our data, and our models. ASEAN becomes a global voice for "Safe AI," influencing international governance.
To build a Sovereign Intelligence Infrastructure that ensures AI serves the long-term flourishing of the ASEAN region.
We are on a mission to equip 1 million citizens with AI safety literacy and to establish Malaysia as the primary "Safe Harbor" for compute governance in Southeast Asia by 2030.
Running offline SLMs in labs.
Equipped with Safety Toolkits.
Signed the 'Safety Oath'.
Digitized National History (Sejarah).
Active R&D pipelines for 2026.
Developing "Safe-by-Design" Small Language Models (SLMs) tailored for the Asian household context. These models run locally, ensuring data privacy while providing educational support without addiction loops.
Target Funder
CoeffectiveGiving + Manifund
Designing standardized, tamper-proof hardware modules for secure AI inference. This project addresses the "Strengthening Security" mandate by creating a root of trust for critical infrastructure.
Target Funder
Institute for Progress (IFP)
Establishing a network of "AI for Science & Safety" nodes in ASEAN. These hubs will focus on Private AI and confidential compute, allowing researchers to collaborate on safety without centralizing power.
Target Funder
Foresight Institute