Across the Gulf, artificial intelligence is moving from glossy strategy documents to daily operations—helping route services, automate decisions, and run systems that people use without necessarily noticing. It is less a single “tech trend” than an enabling layer inside government, business, and infrastructure. The rollout is uneven and still incomplete, but it is no longer theoretical.
In plain terms, “AI” in the Gulf now usually means three things. First is generative AI—systems that can produce text, code, images, and summaries, and are often used to draft, search, or speed up office work. Second is computer vision—software that “sees” through cameras, classifying objects and events, from traffic flows to product sorting. Third is predictive analytics—models that use historical data to forecast risks and optimize operations, such as maintenance schedules, demand patterns, or fraud detection. The Gulf’s most visible gains so far have come from the latter two, where speed and scale matter more than novelty.
Three perspectives show how that shift is taking shape: a young AI engineer navigating opportunity and constraints in Dubai; a UAE-based cleantech company embedding AI into physical recycling infrastructure; and a Saudi AI consultant observing how Vision 2030 and the Humain project are translating national ambition into applied systems. Together, they point to a regional trajectory shaped less by experimentation than by integration, execution, and time-to-deploy—while also raising familiar questions about jobs, data governance, and trust.
For Amr Tamer, an AI engineer at Multiply Media Group and a recent graduate from Khalifa University, working in Dubai is not simply a career choice but a structural advantage. Having lived in Egypt and the UAE, he draws a sharp distinction between talent and opportunity.
I truly don’t think it’s a talent issue
“I truly don’t think it’s a talent issue,” he told The Media Line. “I know a lot of brilliant Egyptian engineers. I know many, many of them, a lot of them are my close friends.”
The constraint, he argues, is systemic. “I think the country is in a bit of an economically difficult situation,” he said about Egypt. “So I’m not sure if they even have the bandwidth right now to have a really deep AI hub or AI culture.”
That difference shapes how young professionals think about risk. “If I lived in Egypt, I don’t think I would be able to be in a situation where I’m financially comfortable,” Tamer said. “The culture is a lot more job-centric; you have to be really focused on just not going broke and not being unemployed.”
Dubai, by contrast, lowers the cost of ambition. “The fact that Dubai is so small, the fact that Dubai has a lot of high-net-worth people, sometimes it puts me in rooms that I would generally maybe not end up in,” he said. “If I just know one person, he knows another person who will help me further.”
That density accelerates professional mobility. “The climb for someone young and hungry like me is a lot easier because the connection is a lot smaller,” he said. “You can find the person you’re looking for a lot quicker than in the US or the UK.”
In a field moving as fast as AI, that speed matters. “We don’t have as much time as we think,” he said. “If you don’t adapt now, it’s kind of getting a little bit too late at this point, and the UAE understood that.”
When Tamer talks about the global balance of power in AI, he comes back to compute, especially graphics processing units, or GPUs, the specialized chips used to train and run many modern AI systems at scale. “The GPU accessibility in the US and China versus the UAE—it’s a very wide difference,” he said, pointing to the sheer concentration of large data centers elsewhere. “There are a bunch of data centers all over the US that hold massive, large-scale GPUs that they can train models on.”
Yet he argues that positioning is not only about size; it is also about adoption and speed. “Other than the really big superpowers, something like the UAE or Saudi Arabia were really the best ones at integrating AI natively,” he said. “They took it in with open arms. They didn’t really fight back.”
Europe, by contrast, hesitated. “Europe was a little bit more apprehensive of integrating AI,” he said. “There was this death-nightmare that the world was going to end, and that slowed things down.”
Give the gift of hope
We practice what we preach:
accurate, fearless journalism. But we can't do it alone.
- On the ground in Gaza, Syria, Israel, Egypt, Pakistan, and more
- Our program trained more than 100 journalists
- Calling out fake news and reporting real facts
- On the ground in Gaza, Syria, Israel, Egypt, Pakistan, and more
- Our program trained more than 100 journalists
- Calling out fake news and reporting real facts
Join us.
Support The Media Line. Save democracy.
For smaller states, he says, centralized direction can mean faster execution. “When you’re a smaller country, you can move faster,” he said. “They have one vision, one thought: ‘How can we maximize the ability for the UAE to be expanded?’”
Tamer also resists the common all-or-nothing framing about jobs. “I think people fall on two sides of the argument,” he said. “Either people say they can do way better than AI in everything, or people say, ‘This thing is going to take my job.’ And honestly, like most things, the answer is somewhere in the middle.”
In his own work, he describes AI less as a replacement than as an accelerant—especially for routine coding tasks such as generating boilerplate, suggesting fixes, writing tests, or turning a rough idea into a quick prototype. “The way I code right now is nothing like I used to code a year ago or two years ago,” he said. “You’re basically trying to balance the workload of 10 engineers into one.”
That compression has consequences. “If 10 engineers become one engineer, then those nine engineers are out of a job,” he said. But he distinguishes between execution and judgment. “What AI displaces is what you do, not who you are,” Tamer said. “It might displace what I’m doing currently, but it cannot displace me as a person and my critical thinking. I can maneuver and adapt and continue to find gaps in what it can do, so I can utilize what it can do for myself.”
Even for power users, the gains are not constant. “AI probably speeds me up 80% of the time and slows me down 20% of the time,” he said, “So, 60% overall is still a net gain.”
One of the clearest fault lines he identifies is between academic training and professional reality. “Coming out of university, I felt like the education was quite weak,” he said. “I felt like the curriculum just hadn’t been updated to the level of what it was.” As a result, competence had to be built independently. “A lot of the things I was learning were outside of the university class,” he said. “It was me studying it on my own and continuing to learn AI, which evolves all the time.”
That emphasis on applied use shows up in physical infrastructure, too. In Abu Dhabi, Sparklo, a UAE-based cleantech company, uses AI-powered reverse vending machines to make recycling more convenient and to feed verified data into recycling supply chains. Maxim Kaplevich, the company’s founder and CEO, says the barrier is still basic access to collection points.
“The core problem is the lack of convenient and engaging collection infrastructure,” Kaplevich told The Media Line. “In the MENA [Middle East and North Africa] region, only around 5 to 7% of plastic is recycled today.”
Sparklo’s machines—Sparklomats—are built around computer vision and verification. A user inserts a bottle or can; the system visually identifies what it is, checks that it meets acceptance criteria, and then records it so the deposit-and-reward loop stays credible for users and auditable for partners. “When a user inserts a plastic bottle or aluminum can into a Sparklomat, the machine identifies the item instantly, accepts it, and credits the user with rewards on the app,” he said.
For us, AI isn’t a headline feature, but the technology that quietly makes the whole experience simple and reliable in real life
“For us, AI isn’t a headline feature, but the technology that quietly makes the whole experience simple and reliable in real life,” Kaplevich said.
He points to both performance and economics: higher recognition accuracy means less contamination and less manual sorting downstream. “Our recognition system identifies the material with around 99% accuracy,” he said. “That results in near-zero contamination and significantly reduces sorting needs and costs.”
Scale is central to the pitch. “Globally, we’ve collected over 200 million plastic bottles and aluminum cans,” Kaplevich said. “That has helped prevent around 30 million tons of CO₂ emissions from entering the environment. Some of the most popular machines in the UAE collect more than 8,000 bottles a day.”
Adoption, he argues, is sustained rather than one-off. “Globally, we now have more than 850,000 users who recycle with us on a regular basis. Many people come with their families,” he said. “Recycling becomes part of a weekly routine.” He links that routine to measurable behavior change: “Around 30% of our users say they started recycling only after discovering Sparklo,” he said.
Kaplevich contrasts Sparklo’s approach with older systems that rely on barcodes, which can fail in the real world. “Most reverse vending machines still depend on barcodes,” he said. “If a label is damaged or missing, the bottle won’t be accepted, and the user ends up frustrated. Our AI-based recognition is designed for real-world conditions, so it can identify bottles even without a barcode or label.”
For companies and municipalities, the value is not only the collected material but also the data trail—counts, types, locations, and performance patterns that can be used for reporting and operational planning. “Users can trust that their actions are counted properly,” Kaplevich said. “And this system is trustworthy for companies too, since the data makes ESG reporting more transparent and measurable, while offering insights on customer behavior and real-world progress.”
Expansion is underway. “Beyond the UAE, we’re actively growing across the region—both in public and private sectors across Morocco, Kuwait, Jordan, and globally. Southeast Asia and CIS countries are on our radar. Only last week we opened in the Philippines,” he said. “In general, we cannot describe Sparklo as an AI project, but more of a sustainable lifestyle platform where AI plays an important role for sure,” he concluded.
In Saudi Arabia, AI is being framed as a strategic national capability aligned with Vision 2030, the Kingdom’s long-term economic transformation plan. A central pillar is Humain, a national initiative to build domestic AI infrastructure, models, and institutional capacity.
Saud Al Dukhyil, an AI consultant and trainer currently working for Prime Source Group, argues that the Kingdom is moving from intent to deployment, especially in domains where governments already control data, systems, and accountability. “Saudi Arabia has moved decisively beyond policy statements into selective and purposeful execution,” he told The Media Line. “Artificial intelligence is no longer treated as an experimental or symbolic layer. In the Kingdom, AI started to be used for transport and logistics, and in urban systems, has been deployed for traffic management and public safety analytics.”
In practice, that can mean AI models that help manage traffic flow and detect incidents, tools that optimize logistics, and analytics that identify patterns quickly across large networks—tasks that are hard to do at speed by human review alone. Still, Al Dukhyil says impact is clearest where governance and ownership are already defined. “The most tangible impact appears where data ownership, scale, and accountability already exist,” he said.
He sees strong interest among young Saudis, but uneven preparation for real deployments. “Youth engagement with AI in Saudi Arabia is high and genuine,” he said. “However, readiness is uneven. The primary challenge is not motivation, but translation. Many young professionals are fluent in using AI tools yet lack exposure to production-grade systems, commercial constraints, and real-world datasets.”
He argues Humain should be judged not by announcements but by whether it changes institutional behavior. “Humain should be understood as a strategic capability-building initiative rather than a branding exercise. If it does not materially change how institutions operate, then it has not fulfilled its purpose,” he said.
For him, the biggest obstacle is not funding or recruitment. “The primary constraint is not talent or capital, but integration,” he said. “Fragmented data environments, inconsistent standards, and limited interoperability across institutions continue to restrict national-scale AI impact.” In other words, even capable models can stall if ministries store data in incompatible systems that cannot easily be combined or used safely.
The current market favors applied intelligence over foundational ambition
Al Dukhyil frames the race as practical rather than philosophical. “The current market favors applied intelligence over foundational ambition,” he said. “This is not a market that rewards generalization. It rewards precision and execution.”
Looking toward the end of the decade, he makes a case for Saudi Arabia as an exporter of applied systems in regulated sectors. “By 2030, Saudi Arabia is well-positioned to emerge as a global reference point for applied AI in industrial and regulated environments. That includes exportable industrial AI platforms, Arabic-native enterprise systems, and governance models that can scale across emerging markets. Thanks to our leadership, there is a consistent goal which will lead Saudi Arabia to be a global benchmark for disciplined and large-scale execution with lasting national impact,” he said.
Even in this execution-focused frame, the familiar risks follow close behind: data privacy, bias in automated decisions, overreliance on systems people do not fully understand, and the labor-market churn that comes when routine tasks compress into fewer roles. The Gulf’s bet, as these three views suggest, is that speed and integration can outpace the disruption—turning AI from a headline into infrastructure.