5,660 Households, One AI Model: Oman Rewrites Its Statistics

On Sunday evening at its Muscat headquarters, Oman's National Centre for Statistics and Information (NCSI) unveiled the results of a project it says is a global first: a Household Income and Expenditure Survey (HIES) built on artificial intelligence rather than the paper-and-clipboard methods statistics agencies have relied on for decades.

📌 Key Takeaways

• NCSI held a launch ceremony on 6 September 2026 to release results from what it calls the world's first AI-powered national Household Income and Expenditure Survey, according to Times of Oman.
• The survey covers a sample of roughly 5,660 Omani and expatriate households across the Sultanate's governorates, the same scale used in NCSI's previous, manually conducted editions of the survey.
• NCSI CEO Dr Khalifa bin Abdullah Al Barwani said the AI approach "saves the time, effort, and resources that traditional methods once required," per Times of Oman.
• The data will be used to reset the weights behind Oman's Consumer Price Index, directly affecting how inflation is measured and reported.
• NCSI first announced plans for an AI-based HIES in September 2025 and confirmed fieldwork was under way by April 2026, according to Muscat Daily and a follow-up Muscat Daily report.

🏛️ What NCSI Actually Announced

The event itself was straightforward: NCSI leadership gathered at the centre's Muscat headquarters to present the completed results of the AI-based HIES, rather than to announce a new pilot or plan. As Times of Oman reported, Dr Al Barwani described the project as "a qualitative shift for Oman" and said the AI-driven data would help "sharpen economic policymaking by providing critical, integrated data on consumer spending patterns" and by "establishing the relative weights needed to calculate the Consumer Price Index."

That CPI link matters more than it might sound. Every few years, statistics agencies have to update the basket of goods and spending weights used to calculate inflation, based on what households are actually buying. That update has traditionally required a full HIES cycle, a labour-intensive exercise involving enumerators visiting households and logging a full year of income and spending. NCSI's stated goal is to compress that cycle using AI-assisted estimation instead.

📊 From Clipboards to Continuous Estimates

NCSI has run the Household Expenditure and Income Survey multiple times since the 1990s using the traditional model, most recently completing a fourth manual edition with the same roughly 5,660-household sample size distributed across the governorates. This new edition keeps that sample scale but changes the underlying method: instead of relying purely on a year-long fieldwork window, NCSI has layered AI models on top of survey data, administrative records and other feeds to estimate spending patterns faster and, it says, more accurately.

The centre first flagged this shift publicly in September 2025, when it said it planned to launch the "world's first" AI-based version of the survey, according to Muscat Daily. By April 2026, Muscat Daily reported that NCSI had reached what it called a "global milestone" in conducting the survey with artificial intelligence, describing the effort as driven entirely by national talent and as the start of a broader roadmap to bring AI into other statistical work at the centre. Sunday's ceremony marked the point at which those results were formally released.

🤝 The Technology Behind the Shift

NCSI's own public statements have focused on outcomes rather than technical architecture, so exactly which models and infrastructure sit behind the new HIES has not been detailed by the centre itself. AiSPRY, a Hyderabad-based AI company, has said publicly that it worked on the engagement: in an announcement from October 2025, the firm said its leadership travelled to Muscat to formally launch the project with NCSI's Statistics Department, in partnership with Otech, the Group ICT arm that Omantel consolidated from Oman Data Park and other technology units. AiSPRY describes the work as applying machine learning to estimate household spending patterns from a mix of survey records, administrative data and utility usage, rather than relying solely on repeated manual interviews, with the system designed to operate within Oman's Personal Data Protection Law and cloud-localisation requirements.

Those details come from the vendor's own account rather than an NCSI confirmation, and readers should treat the specific technical claims as AiSPRY's characterisation of its own work. What is independently verifiable, through NCSI's own ceremony and the CEO's on-record comments, is that the centre has moved from planning to a completed, results-bearing AI system for a flagship national survey, a genuinely rare step for a government statistics office anywhere in the world.

💰 Why the Consumer Price Index Angle Matters

For most residents, the practical effect of this project will show up indirectly, through how inflation is measured. The CPI weights derived from HIES data determine how much influence, say, housing costs or food prices have on the headline inflation figure NCSI publishes each month. If the AI-based survey genuinely produces more current and more granular spending data than the old multi-year fieldwork cycle, Oman's inflation reporting could become more responsive to real shifts in household budgets, useful for everyone from the Central Bank of Oman to businesses setting prices.

It also fits into a broader pattern of NCSI leaning on AI for its statistical work beyond this one survey; the centre has separately run AI-focused public opinion research to gauge how Omanis view the technology, part of what officials describe as embedding AI more widely across the centre's operations.

🇴🇲 Why This Matters for Oman

Government statistics rarely make headlines, but they underpin almost every economic policy decision, from subsidy design to wage negotiations to how the government tracks progress on Vision 2040 targets. If NCSI can genuinely sustain faster, AI-assisted HIES cycles, Oman's policymakers would get more current data to work with than the traditional multi-year survey rhythm allowed. It is also a concrete example, alongside the private-sector AI partnerships expanding around Muscat, of a public institution putting AI into a core operational process rather than a pilot or a slide deck. The real test now is whether the AI-derived weights and indicators hold up against independent scrutiny over the coming CPI cycles, and whether NCSI follows through on its stated plan to extend AI into other statistical fields.