Cisco: 86% of Saudi Organizations Have Hit or Expect To Hit Network Capacity Limits Within Two Years Due to AI

• More than 40% of Saudi organizations have deployed agentic AI enterprise-wide, ahead of the global average of 33%.
• Experts expect local AI network traffic to more than triple within three years, with agentic AI having a particularly high impact
• 86% have already hit or expect to hit campus and branch network capacity limits within 24 months, compared to 73% globally, pointing to the need for continued local network upgrades

New research, released by Cisco in partnership with Foundry, reveals that organizations in Saudi Arabia face a two-year window before network capacity reaches its limit, as AI-driven traffic increases significantly across key AI workloads and attack surfaces expand beyond what current defenses can manage.

The study, based on a survey of 200 IT leaders in Saudi Arabia (and 3,72 across the world), confirms that the rapid rise of large language models (LLMs) and the emerging wave of agentic AI are placing unprecedented strain on enterprise campus and branch networks. In Saudi Arabia, where AI adoption is ahead of global benchmarks, the network is now a major factor in whether enterprise AI deployments succeed or fail.

Tarik Al-Turki, Systems Engineering Director, Cisco Saudi Arabia, said: “The Kingdom of Saudi Arabia is moving at a remarkable pace in adopting AI technologies, with local organizations deploying agentic AI ahead of the global average. But this accelerated innovation brings a critical challenge: network infrastructure must continue to evolve at the same pace. To realize the full potential of AI and align with Saudi Vision 2030, organizations need to prioritize ongoing network modernization as a foundational element of their AI strategy.”

Agentic AI Adoption Is Moving Ahead

More than 40% of Saudi organizations surveyed already have broad, enterprise-wide AI agent deployments, ahead of the global average of 33%. Unlike human users, AI agents trigger dozens of API calls, database lookups, and model inferences in seconds, generating dense east-west traffic that legacy workplace networks were never designed to handle. Saudi respondents report that these AI workloads are more sensitive to network reliability and uptime (73%), packet loss (68%), latency (64%), and bandwidth availability (60%) than traditional applications.

Network Capacity Pressure Is Intensifying as AI Traffic Expected to Triple

Only 29% of Saudi respondents say their networks are fully prepared for projected AI growth (though they are ahead of the global average of 23%). Overall, 86% say they have already hit, or expect to hit, campus and branch network capacity limits within 24 months, compared to 73% globally. Furthermore, Wi-Fi is emerging as a major bottleneck, with 46% of Saudi respondents listing it as the area driving the greatest increase in capacity requirements. Networking and AI experts in Saudi Arabia expect AI’s overall traffic impact to more than triple within the next three years, with the highest individual workload increase, around 116%, attributed to agentic AI.

Despite this growing pressure, 74% of Saudi IT leaders are more confident in their AI strategy than their network’s ability to support it. While nearly 90% plan to modernize their workplace networks, budget remains a barrier. In Saudi Arabia, 34% say budget constraints limit their ability to modernize to a great extent, while 60% say they are limited to some extent.

Attack surfaces are already expanding while observability is a challenge 

AI is also creating a challenging security environment. Most Saudi respondents (93%) are struggling to keep up, 87% say AI is already causing damage, and 64% believe AI-related threats are evolving faster than their security controls. Meanwhile, the observability gap is widening as 58% of Saudi respondents say they lack adequate visibility into AI-related traffic flows across their network.

The findings make it clear that networking resilience, observability, and adaptive security are not supporting acts in the AI era, they are essential for functioning AI. Organizations that treat continued network modernization as a prerequisite to their AI strategy, rather than a parallel workstream, will define the next decade of enterprise AI in Saudi Arabia.