Survey shows 80% of responding companies using AI

PRAGMATIC GROWTH: The analytics company said that companies are turning from finding their way with AI to expanding parts that work for their businesses

• By Crystal Hsu / Staff reporter

• • Add TT as Preferred Source

  • 

  • 

  • 

Taiwanese companies are moving beyond artificial intelligence (AI) experiments to focus on returns from their investments, but gaps in data quality and governance could constrain wider adoption, a report released yesterday by Dun & Bradstreet showed.

More than 80 percent of Taiwanese companies said they have seen returns from AI projects, showing that the technology is beginning to deliver measurable business value rather than remaining at the testing stage, the US business-data provider said.

The Taiwan Corporate AI Momentum Index fell to 60 in the third quarter from 72 in the previous quarter, but remained above the 50-point threshold for expansion in AI investment and adoption. Taiwan’s score was below those of Japan and South Korea, which both posted 66.

An artificial intelligence logo is seen at the Siemens AG booth at a trade fair in Hannover, Germany, on Sunday.

Photo: Bloomberg

The decline does not necessarily signal a pullback in AI spending. Instead, companies appear to be entering a more cautious phase, evaluating which projects generate meaningful returns before scaling them further.

“Taiwan’s market is moving from the rapid adoption of AI to a more pragmatic stage of value validation,” said Julian Prower, Dun & Bradstreet’s executive vice president of global business operations and transformation and general manager for Taiwan.

Nearly 70 percent of respondents have launched AI-related projects, while 30 percent remain in the planning stage, the report said. About 41 percent have reached scaled deployment, including 25 percent that have integrated AI into multiple core business processes and 6 percent that are developing agentic AI workflows, it said.

By contrast, 27 percent remain in the planning or pilot stage, indicating that most surveyed firms have moved beyond proof-of-concept projects and are beginning to deploy AI in actual business processes, it said.

The survey found that 48 percent who responded have seen returns from some AI pilot projects, while 24 percent said multiple projects have generated business value. Another 8 percent reported large-scale and clear benefits, while 1 percent said they have yet to see any return.

The next hurdle is building the data infrastructure needed to scale those gains, the report said. More than half of surveyed companies have yet to fully establish the data foundation required for large-scale AI deployment, highlighting weaknesses in data quality, governance and readiness, it said.

That gap could become increasingly important as companies use AI for decisions such as credit assessments, supply-chain risk evaluation and compliance reviews, Prower said.

“The pace of AI adoption will ultimately depend on the maturity of their data governance,” Prower said. “The challenge is not unique to Taiwan.”

Globally, more than 75 percent of responding companies have reported measurable returns from AI projects, while 28 percent said multiple projects had generated broad or significant benefits, according to Dun & Bradstreet.

For Taiwanese companies, the next phase of AI adoption is likely to depend less on launching more pilots and more on scaling successful applications across the business while maintaining data quality and control.