Artificial intelligence could help Arab governments improve tax collection and make public finances more efficient, but technology alone is unlikely to deliver higher revenues without strong institutions, effective governance and reliable digital infrastructure, according to the Arab Monetary Fund.
The findings were presented during the 17th Ordinary Session of the Council of Arab Finance Ministers in Abu Dhabi, where finance ministers and senior officials discussed fiscal policy, public finances and the growing use of AI in tax administration.
An AMF study covering 114 countries, including 15 Arab economies, examined 570 observations between 2020 and 2024 to assess the relationship between government AI readiness and tax revenue mobilisation.
The analysis found a positive relationship between AI readiness and tax collection. The estimated effect was 0.42 using traditional econometric models and 0.53 under machine-learning models.
The Fund cautioned, however, that greater readiness to use AI does not automatically result in higher tax revenues. Governance quality was identified as the strongest factor influencing tax collection in the machine-learning analysis, alongside regulatory quality, rule of law, trade openness and GDP per capita.
An AMF simulation found that a 10-point improvement in AI readiness could have an average estimated tax-revenue impact of 0.69, although the effect varied significantly between economies.
The estimated impact was 4.63 for oil economies with relatively low taxation, compared with 0.21 for oil-dependent economies. Oil economies with higher taxation recorded a negative reading of 0.7.
UAE Minister of State for Financial Affairs Mohamed bin Hadi Al Hussaini said the country’s experience showed that successful digital transformation required more than investment in technology.
He said the UAE first moved government transactions away from paper before shifting services to mobile platforms and developing digital identity systems. These steps created the infrastructure needed for more advanced AI applications.
The UAE is now examining wider use of AI assistants in government services, including systems capable of answering questions, accessing information and helping users complete tasks.
The AMF highlighted the UAE and Saudi Arabia for using advanced tax tools such as risk-analysis models, compliance controls, electronic invoicing and data analytics. Bahrain, Egypt and Morocco were also cited for digital transformation and AI-related initiatives in tax administration.
About 70 per cent of tax administrations are already using AI, particularly for fraud detection, risk assessment and digital taxpayer services. Global investment in AI applications for tax administrations is expected to reach about $2.3 billion by 2027.
The AMF cited examples from several countries. The United States generated more than $1.3 billion in additional collections through risk-classification models, while Russia recorded €354 million in additional revenue in 2024. France has used satellite imagery to identify undeclared property developments.
The Fund also highlighted challenges including shortages of specialised skills, high implementation costs, cybersecurity concerns, data protection and difficulties connecting AI systems with existing government infrastructure.
The wider fiscal picture remains challenging for Arab governments. Average public expenditure stands at 29.4 per cent of GDP across Arab countries, while current spending accounts for about 79 per cent of total government expenditure.
The AMF said reforms including better-targeted subsidies, direct cash transfers, stronger domestic debt markets and medium-term debt strategies could help governments improve fiscal sustainability alongside digital transformation.
