Singapore leaders split on AI failure accountability

ABBYY has found that business leaders in Singapore are divided over who should be responsible when artificial intelligence systems produce harmful or incorrect results.

Its study found that every surveyed organisation in Singapore now uses AI.

The findings highlight a sharp gap between adoption and confidence. While AI use was universal among Singapore respondents, only 15% said they completely trust the technology to produce accurate outputs.

ABBYY commissioned the research, which Opinium conducted among 1,200 senior managers at companies with more than 100 employees across Singapore, the US, the UK, France, Germany and Australia.

In Singapore, 41% of leaders said responsibility for AI failures should be shared by the organisation and the vendor. That compares with a global figure of 31%, making Singapore the market most inclined towards shared accountability.

Trust also remained limited despite widespread adoption. A net 43% said they trust AI systems to operate without creating unacceptable risks, while 50% trust AI to protect confidential information and 53% trust it to comply with relevant regulations.

Security and compliance concerns featured prominently in the responses. Leaders identified confidential data leakage as the top AI-related security concern, while 83% said they worry about the legal and reputational risks of AI non-compliance, as well as dependence on one or more AI providers.

Human oversight

Singapore stood out in the survey for its reliance on human oversight. Only 6% of organisations said they had no formal approach to keeping people involved in AI decision-making, one of the lowest levels recorded in the research.

That emphasis on supervision coincided with broad use of formal governance structures. Nine in 10 respondents in Singapore said their organisation has an AI governance framework in place, and 85% said governance has made their AI efforts more successful.

Respondents also linked governance to smoother internal execution. Half said it made AI easier to develop, 52% said it helped with scaling, and 60% said it supported adoption across the organisation.

Companies also appear to be examining data and jurisdiction issues more closely. The survey found that 85% of organisations had assessed whether their AI systems could operate across different geographical regions, indicating a degree of maturity on data sovereignty.

Control gap

Even so, the results suggest governance structures are struggling to keep pace with deployment. Nearly half of respondents, 49%, said AI was being adopted faster than their organisation could govern it effectively.

At the same time, 73% said their current governance approach strikes the right balance between mitigating risk and maintaining the speed of AI delivery. That contrast suggests many businesses believe they have the right framework in principle, even as implementation comes under pressure.

Data quality emerged as the biggest obstacle to stronger returns from AI, cited by 20% of Singapore leaders. The findings suggest technical rollout alone is not enough for businesses seeking clearer commercial gains from their AI investments.

The survey also found that understanding of responsibility weakens lower down the corporate structure. While 92% of senior leaders said they know who is responsible for managing AI, that clarity was less evident among junior employees.

This gap in internal understanding may prove significant as companies try to define accountability for systems that are becoming embedded across departments. Clear ownership has become as much a practical issue as a policy one, particularly as businesses weigh legal exposure, reputational risk and supplier dependence.

Roman Kilun, Chief Compliance Officer at ABBYY, commented on the broader findings.

"Businesses are embracing AI, but our research shows that many are still struggling with key questions around accountability, governance and ROI. Senior management need to do a better job of communicating what their responsible AI policies are companywide. In addition, AI requires organizations to know where data comes from, whether it can be trusted, how it can be used, and where it is stored and processed. As data sovereignty grows in importance, a shift toward machine-readable controls that govern data throughout its lifecycle is required to ensure consistent compliance," said Roman Kilun, Chief Compliance Officer, ABBYY.

Kilun also addressed how organisations are moving from trial phases to broader use.

"Organisations we work with are moving beyond pilots and proofs of concept into execution, and clear frameworks and a shared understanding of responsibility are critical to building confidence in AI. At ABBYY, we're committed to being transparent about our own governance approach and supporting customers on that journey," he said.