The conversations around artificial intelligence in the Gulf moved on from exploration to implementation. As governments pursue national strategies and enterprises work on specific use cases, it’s no longer a question of whether AI will be adopted but about how to build the systems that can carry it at scale.
Few developments capture this shift more decisively than Stargate UAE—a vast AI compute project under construction in Abu Dhabi. Led by G42 and supported by a global alliance of technology firms including OpenAI, NVIDIA, Cisco, Oracle and SoftBank, the initiative will see a 1-gigawatt AI supercomputing campus take shape, with 200 megawatts of capacity targeted for deployment by 2026.
“If you look at the UAE’s ambitions around AI, they’re really significant. The country wants to be a global leader in AI. The announcements we’ve seen from the country’s leadership are genuinely commendable. Abu Dhabi, specifically, aims to establish the world’s first AI-powered government. All of this is driven by a clear vision and strategy,” says Abdelilah Nejjari, Managing Director – Cisco Gulf & Levant Region.
“The pace of innovation continues to be incredible,” he adds. “We’re seeing significant developments, especially around AI—preparing customers to adopt and use AI effectively.”
For Cisco, the project signals the region’s readiness to execute on its AI ambitions—and an example of the infrastructure-led transformation the company now sees as central to its role. For Stargate UAE, Cisco is contributing secure networking, observability, and infrastructure across both hardware and software.
“We’re working closely with all of these major players and are very optimistic about the progress so far. We’re proud to be part of something this significant,” says Nejjari.
As part of Cisco’s initiatives, engineers are rethinking how AI infrastructure is managed. Traditional models—where human operators rely on dashboards and alerts—are proving too slow for environments that are increasingly AI-native and real-time. Even AIOps, once viewed as a major breakthrough, is showing its limits.
“Traditional AIOps typically monitor infrastructure and alert you to incidents or issues—sometimes even suggesting a fix,” Nejjari explains. “But they still rely on human intervention. AgenticOps is the next level. It’s fully autonomous, which means the AI agent understands the environment, reasons, makes decisions, and acts—without human input. Multiple agents can coordinate and validate actions at high speed.”
Cisco’s AgenticOps approach was unveiled at Cisco Live 2025, alongside AI Canvas—a unified environment where AI agents and human engineers work together in real time to monitor, diagnose, and resolve issues across distributed environments. The system is powered by Cisco’s Deep Network Model, a domain-specific large language model trained on four decades of telemetry and enriched with over 3,000 reasoning traces from Cisco engineers.
That shift doesn’t eliminate the need for networking engineers, Nejjari emphasises. “As someone who started out as a network engineer, I understand the challenges they face. With AgenticOps, many of those pain points will be reduced—or even eliminated. This allows engineers to focus on higher-value tasks like network design and performance optimisation. The future is bright, and I’m proud that Cisco is leading the way.”
Secure and scalable AI adoption
As enterprises accelerate AI adoption, the demands on infrastructure go beyond scale and speed “To support AI, you need low-latency, high-speed, and resilient networks—especially in this emerging Agentic AI era, where AI agents autonomously make decisions and communicate,” says Nejjari. “Networking also has to be energy-efficient. AI demands a lot of power, so optimising infrastructure helps support sustainability goals as well.”
That level of performance also requires a new approach to security—one that addresses both how AI is used and how it must be protected. Nejjari frames the challenge as a dual mandate: securing the infrastructure that powers AI and using AI to strengthen cyber defence.
“When we talk about ‘AI for security’ and ‘security for AI,’ we mean two things,” he says. “Security for AI means protecting the AI infrastructure itself: the data centres, GPUs, LLMs for training and inference. These systems introduce new risks—hallucinations, prompt injections, and more.”
“On the flip side, AI for security is necessary because attackers are now using AI. Traditional cybersecurity assumes human attackers and human defenders. But AI-powered attacks are faster, more complex, and harder to detect. Human response alone isn’t enough.”
Cisco’s AI Defence platform is built to meet those requirements. It combines runtime protection, intelligent detection, and deep observability with embedded AI agents.
“Cisco’s AI Defence does both: protects AI infrastructure and uses AI in our security solutions,” Nejjari says. “Moreover, we have a unique advantage because we operate at the network layer. Everything passes through the network—so we can embed security deeply.”
Building on that network-layer advantage, Cisco has also developed HyperShield, a system that embeds security enforcement throughout the network fabric. “It’s like placing a firewall at every point in your network. A port on a switch becomes security-aware. AI can determine if a device is genuine or impersonated, and automatically block access if needed. That’s not possible with traditional tools—it has to be AI-driven.”
Human capital in the age of AI
As infrastructure becomes smarter, organisations will need to upskill their teams. The AI shift isn’t simply about new technologies—it’s about new skills. Cisco is investing heavily in regional talent development through its Networking Academy, which has trained over 121,000 learners in the UAE and more than a million across the Gulf.
“The World Economic Forum says 25 percent of jobs will change by 2027 due to AI,” Nejjari says. “Cisco is leading an AI consortium with eight major companies, and we’ve found that 92 percent of IT jobs will be moderately or highly impacted by AI. Most jobs won’t disappear, they will simply evolve. We need to prepare the workforce for that shift.”
To that end, Cisco Networking Academy’s curriculum now includes dedicated AI tracks delivered through more than 40 institutions in the UAE. “Some large government entities in Abu Dhabi have their own academies to train staff,” he says. “The new AI curriculum will empower young professionals and seasoned employees alike to upskill and embrace AI opportunities.”
For all the advances happening at the organisational and national levels, AI also presents deeply human questions—about trust, safety, and the kind of future we’re building. For Nejjari, the implications of AI go beyond business strategy; they carry personal weight as well.
“What concerns me most is the cybersecurity aspect,” he says. “We were already in a constant race with attackers. Now, with AI in their hands, their capabilities will be multiplied a thousandfold.”
He doesn’t sugarcoat the threat. “Attackers only need to be right once. Defenders need to be right every time – AI tips that balance even more in their favour.”
But alongside the concern is genuine optimism. “What excites me most is thinking about my children and the world they’ll live in,” he adds. “The ease and opportunities they’ll have with AI are incredible. Their world will be very different—and I believe, much better.”
“AI is already changing the world dramatically in a short time,” says Nejjari. “And we’re just getting started.”
Discussion about this post