Oracle executive says Asian companies must rethink work to gain from AI

Asian companies risk missing the financial gains from artificial intelligence if they keep adding it to existing processes rather than changing how work is done, Oracle executive Garrett Ilg wrote in a Fortune commentary published June 11. Ilg, Oracle’s executive vice president for Japan and Asia Pacific, said the next phase of enterprise AI will test whether businesses can turn experimentation into operating change.

Ilg said boardrooms across Asia have spent the past three years backing pilots, testing AI assistants and looking for uses across company functions. In his view, executives are now less focused on whether the technology works and more focused on whether it changes the economics of their organizations.

He pointed to McKinsey’s latest global survey, saying it found that companies seeing the strongest bottom-line impact are not just deploying more AI. According to Ilg’s summary of the survey, they are redesigning workflows, governance and decision-making around the technology.

Three stages of adoption

Ilg described the first stage of enterprise AI adoption as assistance. He said AI is already proving useful by giving employees more relevant context, detecting anomalies and helping teams act faster.

Examples he cited include finance teams using AI to find problems before they grow and customer service teams resolving issues more quickly. Those uses, he argued, show value when AI helps workers make better decisions at the moment they need them.

The second stage, according to Ilg, is automation. Traditional automation worked best on repetitive tasks governed by clear rules, while AI can take on more variable and less structured work with reduced manual effort, he wrote.

Ilg said the benefit comes from cutting friction inside companies. Faster approvals and fewer slow handoffs can make businesses more efficient and may change how they grow over time, he wrote.

The third stage is augmentation, which Ilg described as AI expanding what an organization can do. He said AI can help companies coordinate decisions at a scale that would be difficult to manage manually and could support new operating models.

Singapore rail pilot

Ilg cited a project involving Singapore public transportation provider SMRT and Oracle as an example of AI use tied to operations. The companies are piloting JARVIS, an AI-enabled platform meant to combine maintenance and operations data, spot potential issues earlier and help engineering teams intervene before service problems occur, he wrote.

SMRT’s rail network supports more than two million passenger journeys a day, according to Ilg. He presented the project as an example of AI creating value by helping organizations act before operational problems become clear.

Ilg said companies should begin by looking for bottlenecks that cost money or undermine trust, including delays, mistakes, poor handoffs, duplicated work and slow decisions. He wrote that businesses then need to decide what they must change so AI can reduce those problems.

He also argued that governance should help companies use AI in significant decisions rather than block adoption. A business that does not trust its AI systems will be unlikely to rely on them for consequential choices, he wrote.

Ilg’s central warning was that the competitive split will come between companies that build AI into workflows and those that leave it at the edge of the business. Executives needed early experiments to learn what AI could do, he wrote, but the next phase depends on how much organizations are willing to change.

This story draws on original reporting from Fortune.