AI gets to work: F5’s 2025 report reveals enterprises are moving from talk to transformation – CRN

In a striking sign of the times, F5’s 2025 State of Application Strategy (SOAS) Report reveals that artificial intelligence has transitioned from experimentation to enterprise-grade execution. A staggering 96% of global IT decision-makers surveyed are now deploying AI models—an exponential leap from just 25% in 2023. From cost optimization to real-time traffic management and zero-day threat mitigation, AI is now at the core of business-critical operations.

“The report shows IT leaders are gaining confidence in embedding AI into day-to-day operations,” said Lori MacVittie, F5 Distinguished Engineer. “We’re moving toward a future where AI autonomously drives cost savings, efficiency, and security without human intervention. This is AIOps becoming reality.”

AI Gateways: The Next Big Connector

With AI now mission-critical, enterprises are turning to AI gateways to bridge applications with AI tools. Half of the organizations surveyed already use them, while another 40% expect to adopt them within the next year. These gateways serve multiple purposes—protecting models (62%), providing centralized control (55%), and preventing data leaks (55%).

In the Asia-Pacific, China, and Japan (APCJ) region, adoption is accelerating: 49% of respondents are already using AI gateways, with another 46% planning deployment soon. Use cases include traffic observation (61%) and AI demand monitoring, highlighting a shift toward operational AI observability.

Challenges Persist: Security, Costs, and Complexity

However, with this progress come significant hurdles. AI model security tops the list of concerns. Despite enhanced model capabilities, operational readiness gaps linger. Around 60% of organizations still rely on manual workflows, and 54% cite a lack of skilled talent as a bottleneck.

The cost of building and operating AI workloads is rising too—48% called it a barrier in 2025, up from 42% last year. Data governance remains a sticking point, with 39% reporting no scalable data practices and 34% expressing concern over AI output bias and hallucinations.

Another growing pain point: APIs. Over half of respondents (58%) said API sprawl has become an operational burden. Managing complex configurations, vendor APIs, custom scripts, and integrations with ticketing systems consumes precious IT time and introduces friction into automation initiatives.

“Organizations must simplify operations—streamlining APIs, reducing tech sprawl, and automating workflows,” said MacVittie. “Ironically, AI itself can be the answer, as it’s well-suited to manage this complexity.”

Hybrid Cloud: The Dominant Strategy

F5’s findings also reflect an unshakeable shift toward hybrid application environments. An overwhelming 94% of organizations now deploy apps across a mix of public clouds, private clouds, on-premise data centers, edge infrastructure, and colocation sites.

This multicloud strategy offers agility: 91% value the adaptability, while 68% cite improved resiliency and 59% enjoy cost savings. Even for AI workloads, 51% plan to continue using both cloud and on-premises deployments for flexibility and control.

Yet, hybrid isn’t hassle-free. Inconsistent security policies (47%) and fragmented delivery models (53%) are common challenges. Interestingly, 79% of organizations have repatriated at least one workload from the cloud back to on-premise infrastructure—up dramatically from 13% just four years ago. The main reasons? Cost control, security, and operational predictability.

“Despite the complexity, being cloud-agnostic has become essential,” said Cindy Borovick, Director of Market and Competitive Intelligence at F5. “Hybrid deployments are no longer transitional—they’re the default strategy.”

The Road Ahead: Toward Programmable, AI-Driven IT

As enterprises scale AI deployment, F5 emphasizes the need for programmable infrastructure—where systems can dynamically adapt, automate delivery and security, and eliminate dependency on traditional consoles.

By 2026, F5 predicts AI will orchestrate end-to-end IT processes, aided by natural language interfaces and policy-driven automation. This marks a shift toward self-operating IT environments, where AI not only assists but independently governs workflows.

“Flexibility and automation are no longer optional,” said Borovick. “Enterprises that build programmable IT foundations will unlock AI’s full value—scaling transformation while maintaining security, compliance, and user experience.”