This article first appeared in Forum, The Edge Malaysia Weekly on July 13, 2026 - July 19, 2026

The recent Middle East crisis has laid bare a hard truth: countries can no longer treat energy, food and critical industrial supplies as ordinary commodities. Disruptions around the Strait of Hormuz have affected energy and fertiliser flows, linking oil and gas (O&G) security directly to food production and supply costs. For Malaysia, this is not an abstract geopolitical lesson; it is a clear reminder that national resilience begins with the ability to secure, manage and optimise our own critical resources.

Malaysia is blessed with substantial O&G reserves, a deep oil and gas services and equipment (OGSE) capability and a strategic position in Southeast Asia. Petroliam Nasional Bhd (PETRONAS) is widely recognised as a best-in-class steward of these resources, with a long track record of embracing technology to sustain and grow Malaysia’s O&G production. The latest technologies — from real-time data analytics and digital twins to advanced sensing, predictive maintenance and scenario-based planning — empower policymakers and asset owners to not only squeeze more value from existing investments but also to plan new ones more effectively.

By enabling continuous monitoring, risk-aware capital allocation and detailed performance forecasting, these tools help optimise asset uptime, reduce unnecessary expenditures and improve decision-making around field development. This leads to better trade-offs, more accurate project prioritisation and a higher likelihood that new assets will perform as intended. With a target of significantly increasing production of oil equivalent per day, representing substantial growth from current levels, achieving this will require a new level of performance from existing assets and near-flawless stewardship of available capital in the coming years.

At the same time, the country has a new opportunity in data centres. Johor’s emergence as a data centre hub is driven by its proximity to Singapore, strong power and water infrastructure, available land and low-latency connectivity. However, data centres are incredibly energy-intensive. Meeting the rising electricity demand from artificial intelligence (AI) and cloud infrastructure while ensuring adequate energy, food and industrial supplies is no small feat.

This is the country’s challenge: our O&G reserves are valuable but not infinite. Managing them carefully, extending asset life, improving operational excellence and balancing capital investment are problems that even larger and better resourced oil economies have struggled with. The next step is not simply more capital, it is better intelligence.

Most people encounter AI through large language model (LLM) chatbots such as OpenAI’s ChatGPT or productivity tools like Anthropic’s Claude Code. These systems learn from past data and perform well in many scenarios. However, O&G operations behave differently. Wells and reservoirs evolve over time. Feedstocks vary. Catalysts age. Sensor drift occurs. Pumps wear. Compressors trip. A simulator can be accurate but slow, while a machine-learning model can be fast yet may fail outside its training conditions.

The AI that we need is AI that understands physics, engineering constraints, real-time uncertainty and the economic trade-offs of live industrial systems. This is the domain what physics-informed AI, like that of Geminus AI, was built for.

This Palo Alto-based tech company emerged from a University of Michigan research paper and was accelerated by The Hive’s co-creation venture capital studio model. It built a platform that links AI to the physical world by integrating data, physics and computation to continuously understand and optimise complex operations in real time. SLB (formerly Schlumberger) describes physics-informed AI as a way to deliver predictive, real-time optimisation across upstream, midstream and downstream systems, while addressing the limits of traditional O&G operations, such as siloed working, slow optimisation and limited training data.

In practical terms, this means AI that goes beyond pattern-matching in past data — it reasons from the structure of the physical system itself. It can work with physical laws, simulation data, real-time sensors and operational context to produce models fast enough for live decisions, yet grounded enough for engineers to trust.

For Malaysia, this capability could become a national advantage. Our O&G sector is large and mature, where even small percentage gains matter. A 1% or 2% uplift in yield can alter refining economics, and a 2% uplift in offshore production can matter over the life of an asset. Reducing energy consumption across liquefied natural gas (LNG) or petrochemical units can lower operating costs and emissions simultaneously.

Geminus shared several examples at a recent strategic roundtable titled “AI for Energy Sovereignty and Security” in Kuala Lumpur, attended by a broad range of industry participants, where the company showcased its technology and its interest in partnering with Malaysia as a key ally in energy and AI. For instance, in a Bahrain oilfield with about 1,500 wells, the reported result was a 5% production uplift with a three-month payback. In a North Sea upstream operation, the uplift was 10%. In a US Tier 1 refinery fluid catalytic cracker case, Geminus reported reducing model deployment from months to about an hour, enabling optimisation every 15 minutes and achieving a 1.5% yield improvement.

The implication for Malaysia is clear. Physics-native AI can help operators extract more value from existing assets before committing to major new capital expenditure. It can also connect decisions across upstream, midstream, downstream, grids and storage systems, reducing the risk that local optimisations in one part of the value chain create unintended costs elsewhere.

The US is a major source of AI technology, but more importantly for this industry, it is home to next-generation industrial AI companies like Geminus that can optimise critical infrastructure. However, the goal should not be to simply import foreign software. It should be to build Malaysian capability around this technology, and Geminus — in partnership with The Hive, which has a presence in the region — is positioning itself to help make that happen.

Malaysia’s energy data is strategic infrastructure. It contains knowledge about reservoirs, facilities, production behaviour, failure modes, maintenance history, emissions profiles and operational constraints. With the right safeguards, physics-native AI can be calibrated to Malaysian fields, LNG systems, refineries and petrochemical complexes while upholding data sovereignty. Geminus’ positioning is aligned with that principle: customer data remains with the customer, and accumulated implementation expertise helps reduce the learning curve without reusing one customer’s data for another.

While we harness the power of physics-native AI, we are at the same time building reusable industrial AI skills inside Malaysian engineering teams. For instance, Malaysia’s National Energy Transition Roadmap targets net zero by 2050 and is built around energy efficiency, renewable energy, hydrogen, bioenergy, green mobility and carbon capture, utilisation and storage (CCUS). Each of these is a physical system. With AI capabilities that local engineers can use, adapt, validate and scale, the same intelligence that optimises wells and refineries can support carbon capture, power systems, data centres, energy storage and advanced manufacturing.

The future of O&G will not be won by hydrocarbons alone. It will be won by the countries that can operate complex energy systems more intelligently, cleanly and autonomously than others. Malaysia has the assets, talents and urgency. What we now need is the courage to move early and own the intelligence.

Datuk Syed Haizam Jamalullail is managing partner of The Hive Global AI Fund and Greg Fallon is CEO of Geminus AI, a Hive investee US technology company at the forefront of sovereign AI for the energy sector. This opinion piece is part of an ongoing series by The Hive, which explores how private capital drives innovation and growth in Malaysia and the Asean region.

Save by subscribing to us for your print and/or digital copy.

P/S: The Edge is also available on Apple's App Store and Android's Google Play.