UAE Organisations Report Large-Scale AI Agent Deployment, but Containment Remains a Challenge

AI agents are being deployed at significant scale by organisations in the UAE, but new research indicates that many organisations still face challenges in quickly identifying and containing problematic agents across different technology environments.

The research, conducted by Harris Poll for Dataiku across eight markets, found that 62% of UAE CIOs surveyed said their organisations had more than 50 AI agents operating in production. Fifteen per cent reported having more than 500, the highest proportion among the markets included in the study, compared with a global average of 9%.

The survey also found that organisations in the UAE are expanding their AI-related teams. Among respondents whose organisations are creating or planning embedded AI expert or forward-engineering roles, 22% said they expected to recruit more than 10 people over the next year.

The widespread deployment of AI agents has also been accompanied by incidents involving behaviour that did not align with business intent, policies or expectations. Eighty per cent of UAE CIOs surveyed said they had encountered such an incident while the agent remained within its technical operating parameters.

However, 62% said the behaviour had been detected before it resulted in customer, financial, compliance or operational impact. Eighteen per cent reported that such an incident had resulted in actual impact.

“CIOs in the UAE are running agents at a scale few markets have reached, and they’re unusually good at catching bad behaviour before it does damage. But detection and control are different capabilities, and the research shows a real gap between them. An agent you can spot in one system but can’t stop everywhere else isn’t contained — it’s just been noticed,” said Florian Douetteau, CEO and co-founder of Dataiku.

The findings also point to differences between detecting an AI agent behaving unexpectedly and being able to stop it across an organisation. Only 5% of UAE respondents said their organisations could reliably identify and contain a problematic agent across environments within one to two hours. A further 27% said they could intervene later the same day.

For 45% of respondents, containment would take between 24 and 48 hours, while 23% said it would take two days or more.

The challenge is partly linked to the fragmented environments in which AI agents operate. Only 18% of UAE CIOs surveyed said their organisations had a fully unified view of agents across different platforms, data environments and business applications. Twelve per cent said agent lifecycle management was standardised across their organisation.

The research also examined measures organisations are considering to reduce their dependence on individual AI models. Sixty-three per cent of UAE respondents said they planned to use more or different models to support business and AI continuity, while 62% were considering open-source alternatives.

The survey found that AI is also becoming a more prominent issue at board level. All UAE CIOs surveyed said they were experiencing increased pressure from their boards to demonstrate measurable AI returns compared with 2025. Ninety-three per cent said their CEO had indicated that their job security depended on AI outcomes, while 90% said they would personally stake their job on delivering measurable AI results.

The findings indicate that while AI agent deployment is expanding rapidly among UAE organisations, governance, visibility and containment mechanisms are developing at a different pace.

“The UAE has established a strong foundation for agentic AI, with enterprises deploying at scale, investing in specialist expertise, and preventing a high proportion of agent incidents from causing real-world harm,” said Sid Bhatia, Area Vice President and General Manager, Middle East, Turkey, and Africa at Dataiku. “The next stage is to connect those safeguards across the enterprise. That means giving CIOs a unified view of their agent population, clarity on who has the authority to act, and the ability to intervene immediately, regardless of where an agent was built or which system it uses. Rapid containment will be critical to protecting the UAE’s progress as agentic AI becomes embedded across the economy. For CIOs, there is also a significant personal stake in getting this right.”