Technological Independence Relies on an Open Source Foundation

In today’s rapidly consolidating tech landscape, a handful of proprietary providers are increasingly dictating the tools, standards, and rules of engagement for how organizations utilize data and AI. Allowing a few concentrated entities to control the underlying architecture of global technology means letting them decide who participates, at what price, and on what terms.

This comes as AI investments are accelerating across Asia Pacific, with IDC forecasting  AI and generative AI investments in the region to reach USD175 billion by 2028, growing at a compound annual growth rate of 33.6% from 2023 to 2028. As this momentum transitions from experimentation into enterprise systems, the foundation on which AI is built becomes a strategic decision.

In the emerging agentic economy, business decision-makers cannot afford to mistake a rented capability for permanent operational control. Many enterprise AI strategies are currently putting the cart before the horse by focusing entirely on the model layer. We treat generative tools as if AI begins and ends with a prompt. It does not. In the rush to deploy these technologies, organizations are focusing entirely on the model layer and overlooking a fundamental truth: AI begins below the model, with data, architecture, and governance.

If those foundations are concentrated, opaque, and inaccessible, the AI built on top will reproduce those exact conditions, only faster, and fail at a far greater scale.

Open source is the ultimate defense against vendor lock-in and obstacles to innovation. Instead of dependence on a single supplier, organizations can participate in a shared, democratic ecosystem.

To build dependable, production-grade enterprise systems rather than impressive demonstrations, leaders must realize that true technological independence relies on three pillars: interoperability, sovereignty, and private AI.

1. Interoperability is freedom of choice

Interoperability is essential to a business’s data strategy to prevent permanent vendor lock-in. Interoperability allows an institution to replace one component without changing everything. Or without forcing all of their data into one single place or vendor.

An open model running on proprietary data formats, proprietary orchestration, and proprietary cloud interfaces is still a locked system. When any single provider can unilaterally condition access or change pricing, operational agility is compromised.

A hospital should be able to use one storage system, another analytics engine, and a locally adapted AI model without copying sensitive patient data into a proprietary island. A government should be able to replace a technology provider without having to rebuild an essential public service. A researcher anywhere should be able to use open formats and interfaces and adapt them to local languages, local data, and local needs. Seamlessly.

Adopting open table formats like Iceberg, open engines and catalogs, open source compute engines and APIs, and common governance policies allow multi-disciplinary teams to work on the same governed data without forcing unnecessary copies or moving all of the data to a single provider for it to deliver value. Needless to say, this will be more cost-efficient. At the model layer, Singapore’s SEA-LION shows how open-source AI can be adapted for local needs as a Southeast Asia-focused large language model (LLM) that is designed to reflect the region’s linguistic and cultural context.

With growing volumes of structured and unstructured data, organizations need the flexibility to choose and run any model in any environment where the data resides. This fundamental decision prevents today’s procurement decision from becoming tomorrow’s permanent dependency.

2. Sovereignty is a condition for business continuity

Digital sovereignty is no longer just a policy preference for highly regulated sectors but is emerging as an operational requirement for any resilient enterprise.

AI is becoming a fundamental building block of public administration, healthcare, education, defence, finance and more. In fact, the technology itself becomes the critical infrastructure. As AI integrates into core operations, maintaining control over that technology determines continuity.

Relying entirely on external public clouds or proprietary black boxes introduces significant operational risk. Third-party providers can abruptly alter token pricing, modify license terms, deprecate models, or adjust terms of service. Geopolitical shifts and compliance mandates can restrict access overnight.

Organizations and entire countries are building processes around capabilities that they do not ultimately control. Whether the decision originates in a boardroom, a regulator’s office, or a geopolitical confrontation, the operational result can be the same: access disappears.

And an institution that believed it had bought a technological capability suddenly discovers that it had only rented permission to use it.

An AI sovereign strategy must answer the following questions:

1. Where does our data reside?

2. Who can access it and how?

3. Which models can use it?

4. Can we move the workloads?

5. Can we replace the models instantly?

6. Can we audit and inspect the system?

7. And most importantly: Can we continue operating if a provider changes its commercial or political position?

Sovereignty does not mean isolation. It means being able to participate in a global ecosystem without surrendering control of essential capabilities. This entails having the hybrid capability to execute workloads seamlessly across public clouds, private data centers, or at the edge, ensuring your operational roadmap remains entirely under your jurisdiction.

3. Private AI is a condition for operational  responsibility

IDC notes that only 7% of enterprises in Asia Pacific are highly prepared in terms of governance, risk and compliance (GRC) capabilities to support new AI and data regulations. This readiness gap makes Private AI an operational necessity. To extract competitive advantage from proprietary data without exposing intellectual property, enterprises must adopt a Private AI framework: Open in its foundations, private in its operation, and accountable in its outcomes.

Instead of continually moving massive datasets to external AI applications, businesses must bring compute directly to the governed data. Running models within a verifiable, trusted perimeter allows organizations to apply consistent security, identity access management, and data lineage controls across the entire lifecycle.

Participation in open source projects as well as building our platform on open data standards ensures that enterprise data and AI infrastructure remain flexible, portable, and entirely under the organization's control.

Moving beyond the demonstration

The market is saturated with impressive technology demonstrations. However, enterprise value is generated by resilient systems that work predictably on Monday morning, protect sensitive assets, and pass stringent audits on Friday afternoon.

Responsible deployment requires identity and access controls, data lineage, evaluation, bias and security testing, read-teaming tools, continuous monitoring, transparent incident reporting, audit logs, human oversight, appeal mechanisms, and the ability to shut a system down.

All that should become digital public goods. This shared safety layer is one of, if not the most, crucial pieces in today’s AI ecosystem.

By anchoring AI strategies in hybrid flexibility, open data architectures, and sovereign control, business leaders can build a scalable foundation that drives measurable performance while mitigating risk. We believe that open source will not merely make AI cheaper. It will make AI more contestable, adaptable, locally relevant, and worthy of trust.

The views and opinions expressed in this article are those of the author and do not necessarily reflect those of CDOTrends. Image credit: iStockphoto/Boonyachoat