CIOs shoulder AI workforce redesign as governance lags

Thoughtworks has published research showing that Chief Information Officers are taking on more responsibility for workforce redesign than for core IT infrastructure. The survey also found that organisations in Singapore are using a more distributed approach to AI decision-making.

The study surveyed 3,200 CIOs across 10 countries and examined how AI is changing technology leadership, governance and accountability. Globally, 89% of respondents said they are now more responsible for redesigning workforce workflows and labour models than for managing core IT infrastructure. In Singapore, 84% said the same.

At the same time, the research suggests many companies are struggling to align governance structures with the speed of AI adoption. Globally, 88% of CIOs said AI adoption in their organisation is moving faster than governance can adapt, compared with 80% in Singapore.

Preparedness levels in Singapore were mixed. While 36% of respondents said their organisation was very prepared to govern AI consistently across business functions, 58% said they were somewhat prepared.

Visibility into AI tools adopted outside central technology teams was also lower in Singapore than globally. Some 78% of Singapore CIOs reported full or high visibility into AI tools or workflows adopted independently by business units, compared with 89% globally. A further 21% in Singapore reported moderate visibility.

Distributed control

The findings suggest there is no settled model for who leads enterprise AI. Globally, 23% of respondents said the Chief Executive Officer has the greatest influence over AI decisions, followed by central IT or technology leadership at 21%, the executive leadership team at 11% and dedicated AI roles at 10%.

Budget ownership was similarly dispersed. Some 22% said AI budgets are managed centrally by IT, 22% said responsibility is shared between IT and business units, 20% said budgets are controlled independently by business units, 19% said they are managed at executive or board level, and 17% said the model is still evolving.

According to the survey, that distribution creates a gap between authority and accountability. Nine in 10 CIOs globally said central IT would still ultimately be held responsible for security breaches or compliance failures caused by AI tools purchased independently by business units. In Singapore, 80% said central IT would still be held responsible for such failures.

CIOs also reported personal exposure to risks they do not fully control. Globally, 37% said they feel personally accountable for security incidents involving AI systems, 35% cited data privacy breaches, and 34% pointed to brand or reputational damage from AI misuse.

Another 35% said they feel personally accountable for workforce disruption caused by AI adoption despite not being able to fully influence the outcome.

"AI governance is also a workforce design issue," said Rachel Laycock, Chief Technology Officer, Thoughtworks.

"As AI changes how work gets done, organisations need to rethink roles, workflows and decision rights so people know where human judgment is still essential and where AI can take on more of the work. Training matters, but it's only one part of building an organisation that can use AI effectively at scale."

Leadership model

The study also examined the rise of the Chief AI Officer. It found that 70% of organisations surveyed have already hired a Chief AI Officer, while a further 26% are looking to do so. In Singapore, 77% said their organisation has already hired one.

Even so, the relationship between that role and the CIO remains unsettled. Some 36% globally said the Chief AI Officer acts as an extension of the CIO's central strategy, while 35% said the role operates independently with equal or greater influence across the business. Another 29% described the relationship between the CIO and Chief AI Officer as a source of organisational friction or unclear boundaries. In Singapore, 23% said that relationship creates friction or unclear boundaries.

Thoughtworks executives said AI oversight cannot be treated separately from broader data and cost questions. The findings point to AI governance as an issue that cuts across business units, with data ownership, system selection and spending decisions often sitting with different leaders.

"AI governance can't sit apart from data governance or from the economics of AI use," said Shayan Mohanty, Chief Data and AI Officer, Thoughtworks. "The person accountable for the data may not own the AI systems using it, while the people choosing those systems increasingly sit across the business. As adoption scales, organisations need the visibility and governance to understand where AI is creating value and where cost and risk are accumulating."

In Singapore, the response appears to include a focus on skills. Among actions taken or planned, 30% of respondents selected upskilling technology staff and 20% selected upskilling the wider workforce.

Thoughtworks' own Chief Information Officer argued that the issue is no longer about assigning ownership to one executive. "At Thoughtworks, our experience has been that AI transformation is a team sport, from defining enterprise AI strategy and architecture to embedding AI into internal platforms and day-to-day operations," said Xia Jie Jessie, Chief Information Officer, Thoughtworks. "The question isn't who owns AI, but how leadership collaborates to create business value responsibly and at scale."

"Authority over AI is distributed, but accountability hasn't always moved with it," said Mike Sutcliff, Chief Executive Officer, Thoughtworks. "The answer isn't to pull every decision back into central IT or put one executive in charge and assume the problem is solved. Organisations need clearer decision rights, and people need the skills and information to make good decisions as AI becomes part of how the business runs."