90% of leaders eye agentic AI, only half understand the risks

C-suite executives remain confident in their artificial intelligence (AI) systems, despite many organizations having limited governance controls in place to manage risks effectively.

This is according to the 2025 EY Responsible AI Pulse survey, which was conducted between March and April 2025. Respondents gathered the views of 975 C-suite leaders across 21 countries, including 30 from Singapore.

All (100%) of Singapore respondents (global 99%) said their organizations have already integrated AI across most or all initiatives or are actively doing so as part of a broader strategic plan. 

Yet, only 53% (global 71%) have moderate to strong controls to protect AI systems from threats such as unauthorized access, corruption or theft.

However, this confidence contrasts with the findings of the earlier EY Reimagining Industry Futures study conducted in November 2024, where only 30% of the respondents in Singapore (global 49%) reported they have either fully integrated or selectively rolled out AI into critical business and IT workflows across their organizations. 

The majority (Singapore 70%, global 51%) were still at pilot or proof-of-concept stages. Singapore respondents of the EY Reimagining Industry Futures study included C-suite leaders and senior decision makers like operations directors and product heads. The differing perspectives may reflect a gap in how executives and other management personnel interpret what fully integrating AI means in practice.

“There is clear ambition to scale AI across organizations, but ambition must be grounded in operational reality,” said Manik Bhandari, EY Asean Artificial Intelligence and Data leader.

“True integration requires reengineering core business processes and redesigning functional roles,” he said. “With agentic AI, there may be a complete rewiring of workflows. This in turn, reshapes workforce structures.”

In Singapore, the government has taken proactive steps to support enterprises on their AI journey through programs like the Enterprise Compute Initiative. Even with this strong ecosystem, organizations need alignment between business and technology leaders. Otherwise, many may risk overestimating their progress.”

The EY Responsible AI Pulse survey also pointed out that while C-suite executives are invested in the technology’s potential, close to half (Singapore 47%, global 51%) admit it is challenging to develop governance frameworks for current AI technologies. 

More concerning is the outlook for governance of the next generation of AI technologies. 67% (global 49%) agree their organization’s approach to technology-related risks is insufficient to address new challenges.

Despite the growing use of AI, fewer than half (Singapore 43%, global 50%) are actively investing in governance frameworks to address the risks and challenges of these emerging AI technologies.

Majority of C-suite executives expect to use emerging AI technologies within the next two years. While 90% of executives in Singapore (global 94%) are currently using or planning to use agentic AI in the next two years, only 55% (global 56%) are familiar with the associated risks. 

Even bigger gaps exist across other emerging technologies like multi-modal AI, where 86% (global 94%) are utilizing or planning to utilize the technology in the next two years, but only 43% (global 54%) are aware of risks.

- Advertisement -