CIOs seek self managing networks

Enterprise CIOs are shifting their priorities, no longer wanting to spend time keeping networks running, but instead seeking infrastructure that can manage itself. According to Sajan Paul, General Manager, HPE Networking India, this change is driven by the increasing importance of artificial intelligence in business transformation.

CIOs are now measuring technology investments by the business outcomes they enable, rather than the infrastructure they maintain, placing fresh demands on enterprise networks to become intelligent, autonomous, and resilient.

“The CIO’s agenda is no longer to keep the network alive. Their agenda is to ask what AI projects will benefit the business,” Paul said. “Their expectation today is, ‘I wish my network is self-driving so I can focus on higher-order business problems.'” The urgency stems from the rapid pace at which AI is reshaping enterprise infrastructure.

Related: Dabur partners with Accenture to scale AI

Networking technologies that once evolved over several years are now advancing within months, as AI workloads demand exponentially higher bandwidth, lower latency, and real-time communication across distributed environments. Historically, enterprises migrated from 1 Gigabit to 10 Gigabit, then to 40 and 100 Gigabit networking over multiple technology cycles.

Today, the industry has already moved towards 400-gigabit, 800-gigabit, and even 1.6-terabit switching as AI infrastructure scales rapidly. “It’s not innovation for innovation’s sake. AI workloads are demanding it,” Paul said, adding that networking has become the foundational layer determining whether enterprise AI can scale successfully.

The changing setting is altering how enterprise networks are expected to operate. Traditionally, networking teams have relied heavily on human intervention for monitoring, troubleshooting, and performance optimization. But as AI applications spread across cloud, edge, and enterprise environments, Paul believes that operating model is becoming increasingly difficult to sustain.

Instead, enterprises are moving towards what Gartner terms “agentic network operations”, where AI continuously monitors network behavior, diagnoses issues, recommends corrective actions, and progressively automates routine operational decisions. “Automation is the traditional word. We need AI-led automation, what we call a self-driving network,” Paul said.

Related: CISA merges lessons into vulnerability disclosure guidance

Building such autonomous capabilities requires years of investment in training AI models using millions of historical networking incidents and technical support cases, rather than relying on generic public AI models. The shift, however, is not merely about preparing networks for AI workloads. Increasingly, AI itself is expected to operate the network, making infrastructure capable of learning from previous incidents, predicting failures, and improving operational efficiency over time.

Paul also argued that enterprises looking to scale AI can no longer rely on fragmented networking environments assembled from multiple vendors. “If you need to harness the potential of AI, we need to have a purpose-built network, or AI-native networking,” he said. For HPE, AI-native networking means AI becomes part of every critical decision within the network, rather than being added later through dashboards or monitoring tools.

“AI should be fundamental to every decision the network takes, not an afterthought.” Looking ahead, Paul expects India’s emerging AI infrastructure ecosystem to create an entirely new category of networking demand, with AI factories, GPU clusters, and hyperscale data centers expected to be distributed across multiple locations.

Related: Study reveals hard drive reliability trends

As AI becomes embedded across enterprise infrastructure, networking is also emerging as a key aspect of enterprise operations. Paul believes identity, machine learning, and networking will need to converge as organizations deploy connected devices, sensors, and edge systems. “In the network, the first packet is seen there itself. That’s why every network device has to be designed with built-in security,” he said.

Validating the identity of users, devices, and sensors will increasingly depend on AI and machine learning. For Paul, the larger transformation extends beyond networking technology to the evolving role of enterprise IT leadership. As AI adoption accelerates, CIOs are spending less time managing infrastructure and more time driving business growth, customer experience, and competitive differentiation.

Infrastructure, in turn, will increasingly be expected to manage, optimize, and heal itself. “The conversation has shifted,” he said. “CIOs want to focus on customer experience and business outcomes. The expectation now is that the network should increasingly take care of itself.”