(Yicai) Feb. 27 -- While businesses in the Asia-Pacific region plan to hike investment in artificial intelligence this year, nearly all of them are unprepared to handle the associated risks, according to recent reports from Accenture.

Eighty-seven percent of APAC business leaders plan to raise investment in AI in 2025, the Accenture Pulse of Change Survey showed, but only 1 percent are fully prepared to handle the risks, revealing a critical gap between ambition and readiness, according to the report Responsible AI: From Compliance to Confidence.

The professional services company conducted the survey between last November and December, gathering insights from 790 C-suite executives and 687 non-C-suite employees across 22 industries and five countries in the APAC region, including China. The latter report was based on more than 200 respondents in that area.  

“Only 1 percent of APAC companies feel confident in their ability to roll out responsible AI at scale today,” Vivek Luthra, who runs Accenture's regional data and AI business, noted yesterday.

Further analysis shows that APAC businesses lag in generative AI investments and adoption. The region has the lowest generative AI spending growth compared with others globally. And less than half of APAC executives anticipate extensive adoption of genAI across business units. Fewer still anticipate enterprise-wide adoption.

The region faces unique infrastructure challenges that impede AI adoption, according to Senthil Ramani, APAC lead for data and AI at Accenture. “While many clients have great aspirations and are ready to scale AI impact in their businesses, they often face the hurdle of legacy digital core systems that need modernization,” he said.

He cited Thailand's banking sector as an example, where lenders eager to implement AI-driven customer experiences find themselves constrained by outdated backend systems still running on mainframes.

“The fundamental premise we've observed across clients globally is the importance of having the right AI scaffolding in place,” Ramani pointed out. "This includes having data on the cloud, properly structured data estates with correct lineage, appropriate governance, and basic principles of responsible AI implemented."

Talent Gap 

Companies already extracting substantial enterprise-level value from generative AI are six times more likely to significantly increase genAI investments this year, positioning themselves to compound their advantage, according to the Pulse of Change Survey.

In China, for instance, 64 percent of businesses are mainly investing in AI so as to develop new AI-powered products and services, while 28 percent are focused on improving existing processes and workflows.

This tech-centric approach may be creating critical gaps. Chinese companies are allocating 57 percent of their AI spending to core technology infrastructure and data, compared with just 43 percent for talent development.

This imbalance manifests in perception gaps between management and staff. While 96 percent of C-suite executives believe their workforce has received adequate AI training, only 78 percent of employees agree. Among those staff currently not using AI tools, 38 percent cite lack of training resources as the primary barrier, followed by concerns about output accuracy (35 percent) and unclear ethical guidelines (15 percent).

Third Pillar

Apart from technology and talent, the Responsible AI report identifies trust as the third essential pillar for successful AI implementation. Of the APAC business leaders surveyed, 78 percent acknowledged that AI's full potential depends on establishing a strong foundation of trust, while 48 percent view responsible AI as a revenue growth driver.

“As businesses across APAC deal with change and disruption, they recognize that success lies in embracing flexibility and finding new sources of efficiency through technology,” said Ryoji Sekido, chief executive of Accenture Asia Oceania.

“They have increased their investments in AI, but the majority are finding it difficult to extract the right value from this investment,” he said, adding that organizations need a three-pronged approach to succeed.

“To effectively scale AI, particularly generative and agentic AI, businesses need to invest in building trust among their people and customers, ensure they have the right data foundation, and operationalize responsible AI,” Sekido noted. “That's the only way of creating long-term, sustainable value.”

Editor: Tom Litting