THE RISE OF THE AUTONOMOUS ECONOMY: A 2025 RETROSPECTIVE FROM THE MIDDLE EAST

Kayvan Karim, Assistant Professor at School of Mathematical and Computer Sciences, Heriot-Watt University Dubai

The year 2025 will likely be remembered as the moment the global economy stopped simply automating tasks and started handing over the keys to autonomous agents. For decades, the promise of automation was simple: machines doing repetitive work faster than humans. But the last twelve months have ushered in a fundamental paradigm shift. We have moved from the era of static scripts to the age of “Agentic AI”, systems that don’t just follow orders but perceive, reason, and act to achieve complex goals.

In their 2025 Technology Trends report, Accenture’s analysts have termed the explosion of these capabilities as “The Binary Big Bang”. As generative AI becomes central to enterprise technology, the cost of development has plummeted, leading to a proliferation of new systems where digital agents act autonomously. These systems have given rise to Agentic AI, which acts as a proactive partner rather than a passive interface. These agents are now capable of “Superagency,” a collaboration architecture that orchestrates multi-agent systems to handle complex workflows that require specialised knowledge across different domains.

This shift is nowhere more palpable than in the Middle East. From the giga-projects of Saudi Arabia to the smart logistics hubs of Dubai, the region is leveraging this technological inflection point to decouple its economic future from hydrocarbons and rebuild it on a foundation of silicon and code.

The Economics of Intelligence

The catalyst for this revolution is a dramatic collapse in the cost of cognitive labour. When examining the economics of intelligence, Stanford University reported in its 2025 AI Index Report that the catalyst for this explosion in autonomy is the radical democratisation of computing power. Between late 2022 and late 2024, the inference cost for a system performing at the level of GPT-3.5 dropped over 280-fold. This trend accelerated through 2025, with hardware costs declining by approximately 30% annually and energy efficiency improving by 40% each year.

These economic shifts have lowered the barriers to entry, moving advanced AI from the realm of massive research labs to the operational budgets of mid-sized enterprises. As Menlo Ventures noted in their mid-year update, enterprise spending on model APIs more than doubled to $8.4 billion in the first half of 2025 alone, signalling a decisive shift from experimental “training” budgets to production-grade “inference” budgets.

The Middle East’s Sovereign Pivot

In the Gulf Cooperation Council (GCC), this technological wave is being ridden with strategic intent. The region is not content to merely import Western or Eastern models; it is building its own “Sovereign AI.”

In the UAE, the Technology Innovation Institute (TII) has continued to push boundaries with its Falcon series. As highlighted by ITU in 2025, the Falcon LLM has evolved into a multi-modal framework capable of processing vision and audio, enabling it to interpret complex documents and charts locally without data leaving the country. Similarly, G42’s Inception has solidified Jais’s position as the world’s premier Arabic-centric model. By integrating Jais into the Microsoft Azure Model Catalogue, they have provided generative AI access to over 400 million Arabic speakers, ensuring that the nuances of the region’s language and culture are preserved in the digital age.

Saudi Arabia has matched this ambition with the launch of Humain, a PIF-backed AI champion. According to Reuters reports from late 2025, Humain is not only building massive data centre capacity but is also developing a voice-first operating system designed to replace traditional icon-based interfaces. This aligns with the Kingdom’s broader Vision 2030 goals, where AI is expected to contribute over $135 billion to the economy.

From Automation to Autonomy in Industry

The distinction between “automation” (following rules) and “autonomy” (making decisions) is best illustrated in the region’s critical infrastructure.

In the energy sector, Saudi Aramco and Yokogawa achieved a historic milestone at the Fadhili Gas Plant. As reported by Oilfield Technology, they successfully deployed autonomous control AI agents that utilise reinforcement learning to optimise the Acid Gas Removal unit actively. Unlike traditional systems, these agents adapt to changing environmental conditions without human intervention, reducing chemical and steam consumption by up to 15%.

Similarly, ADNOC partnered with G42 and Microsoft to launch “EnergyAI.” This agentic system automates complex tasks such as seismic analysis and geological modelling, compressing workflows that used to take months into mere days.

In logistics, the shift is physical. DP World has revolutionised container handling at Jebel Ali with the BoxBay system. As described by Marine Insight, this high-bay storage technology stacks containers up to 11 tiers high in a steel rack, allowing fully automated cranes to access any container without having to reshuffle others. This change increases terminal capacity by 300% and creates a safer, more efficient operating environment.

The GenAI Divide: Enterprises vs. SMEs

While giants like Aramco and DP World forge ahead, the picture for Small and Medium Enterprises (SMEs) is more complex. Project NANDA’s 2025 research highlights a “GenAI Divide,” revealing that while 95% of organisations are investing in AI, only 5% are extracting significant value.

For SMEs, the barriers are talent and infrastructure. However, the rise of Low-Code/No-Code platforms is providing a bridge. As reported by Gulf News, Zoho has seen 50% growth in the region, driven by businesses modernising legacy systems without the need for expensive engineering teams.

To further support this sector, the Saudi SME Bank launched Phase II of its Agency Model in 2025. By partnering with crowdfunding platforms like Manafa and Lendo, they have allocated SAR 240 million specifically to finance SME growth and digital transformation.

The Future of Work: A Divergent Path

The impact on the job market is profound. The World Economic Forum’s “Future of Jobs 2025” report predicts a divergent effect: while routine roles in administration and manual labour are declining, demand for AI and big data specialists is surging.

In the GCC, this dynamic intersects with nationalisation agendas. Governments are using AI to solve the skills mismatch. The Massar Al Ghurair platform, launched in the UAE in 2025, uses AI algorithms to match Emirati youth with career paths and upskilling opportunities. By automating career counselling and recruitment, the region aims to replace low-skilled expatriate labour with high-skilled local talent.

Looking Ahead to 2026

As we look toward 2026, the focus will shift from adoption to governance and integration. Gartner forecasts that IT spending in the MENA region will reach $169 billion in 2026, an 8.9% increase mainly driven by AI infrastructure.

We can expect the realisation of “Cognitive Cities.” In Saudi Arabia, NEOM is moving from earthworks to deploying a cognitive operating system that predicts resident needs. Meanwhile, Dubai’s Cashless Strategy aims to have 90% of all transactions be digital by 2026, creating a data-rich environment for further autonomous innovation.

The year 2025 was the year the machines started to think. The year 2026 will be the year we learn to live and work alongside them.