Ranvish Vir, Growth Lead for Middle East and APAC at Nebius

As artificial intelligence (AI) moves from experimentation to enterprise adoption, 2025 is set to be a transformative year, particularly in Southeast Asia (SEA).

In an interview with e27, Ranvish Vir, Growth Lead for Middle East and APAC at Nebius, outlined the key trends likely to shape the region, from infrastructure scaling to sovereign models and industry-specific adoption.

One of the most significant shifts expected next year is the rise in demand for AI-optimised cloud and compute infrastructure. As more companies embed AI into core business operations, existing systems are being stretched to meet growing needs for real-time processing, model fine-tuning, and large-scale inferencing.

“We expect cloud and compute power demand in SEA to accelerate significantly in 2025,” said Vir. “As AI infrastructure investment surges, led by companies like Nebius, Nvidia, and Accel-backed cloud providers, businesses will increasingly turn to AI-native cloud platforms and GPU-powered infrastructure to scale their AI capabilities.”

Traditional general-purpose cloud platforms are giving way to infrastructure explicitly designed for the specific workloads. Vir pointed out that the need for scalability and cost efficiency propels this shift.

Also Read: How Agentic AI will create telecom’s first truly autonomous workforce by 2030

Several sectors stand out as early and aggressive adopters of AI technologies in 2025.

“In 2025, the most aggressive adopters of AI technologies in Asia will be fintech, healthcare, logistics/supply chain management, gaming, and robotics,” said Vir. “Each industry is driven by efficiency gains, cost optimisation, and regulatory or consumer-driven advancements.”

Fintech is expected to see increased use of the technology in fraud detection, algorithmic trading and personalisation of digital banking. “AI adoption … will accelerate as financial institutions seek real-time risk assessment, personalised financial services, and enhanced cybersecurity,” he added.

Vir further highlighted how functionalities vary across industries. For example, healthcare is becoming a key beneficiary of predictive analytics, diagnostics, and accelerated drug discovery.

“AI-driven predictive analytics, autonomous decision-making, and real-time optimisation will be essential for managing complex logistics networks, last-mile delivery, and inventory forecasting,” he noted.

“The gaming industry is also emerging as a key frontier, with developers using generative models and agentic AI to power more adaptive, immersive, and personalised player experiences,” he said. “At the same time, robotics is gaining traction across manufacturing, healthcare, and smart city deployments.”

Rise of sovereign AI and localised models

Another defining trend for 2025 will be the rise of sovereign AI initiatives and developing homegrown models by startups in Asia.

“The next wave of AI competition will be defined by how well homegrown models adapt to regional business needs while ensuring AI sovereignty and data security,” Vir explained. Local models offer a strategic advantage as AI laws tighten and privacy regulations evolve.

Vir stated, “The increased availability of AI infrastructure can pave the way for more homegrown AI models spearheaded by Asian startups. These have the potential to rival global players in performance and availability.”

Also Read: The product management strategy behind building AI agent platform

Examples from countries such as Singapore, India, and Indonesia—each investing in national AI strategies—underscore this shift. “We expect to see a rise in sovereign and national AI models … with governments providing funding, regulatory support, and sandboxes for startups to test new models,” he added.

Despite the momentum, several challenges could slow progress. First among these is regulatory complexity, particularly around data localisation.

“Cross-border data transfers will become more restricted, creating compliance challenges for businesses relying on global AI models and cloud services,” said Vir. “Countries enforcing stricter data sovereignty regulations will require companies to rethink data storage, processing, and AI deployment strategies.”

Talent remains another significant bottleneck. “The gap between AI talent demand and supply will become more pronounced,” he warned. Citing a recent survey, Vir noted that “one in three Singapore businesses struggle to find AI talent in the country.” This could increase the use of AI-assisted software development and low-code/no-code tools to compensate.

Energy consumption and sustainability are growing concerns, especially as governments push for greener infrastructure. “Energy consumption concerns related to AI workloads will become more pressing,” he added.

Geopolitical risks also loom large. “Increasingly complex export control regulations and geopolitical considerations … can restrict access to state-of-the-art GPUs and AI hardware in certain markets,” said Vir.

As businesses and governments across SEA double down on investment, 2025 will be marked by a race to scale responsibly.

“Infrastructure providers, startups, and governments must work together to create sustainable, high-performance AI ecosystems tailored to regional needs,” said Vir.

—

Image Credit: Nebius

The post 2025 trend forecast: Fintech, healthcare, and robotics lead the AI charge in Asia appeared first on e27.