AI possibilities and readiness in enterprise networks

Not long ago, the idea of asking AI a question was a clever novelty. But today, AI is everywhere. It’s serious, robust, game-changing—including in enterprise network management, where the adoption of artificial intelligence/machine learning surged 40% over the past year, according to a recent Network World article. Network World showed that, in the last 11 months of 2024, a total of 3,624 terabytes of data was transferred to and from over 800 AI applications such as ChatGPT.

But are enterprise network decision-makers implementing AI the right way to drive measurable benefits? According to a new report by MIT’s Project Nanda, 95% of enterprise AI projects are failing, leading to no measurable financial return. Despite billions in investment, most projects fail to move beyond the experimental stage, creating a "Gen AI divide" between a few high-value successes and widespread stagnation. The problem is not the technology, it’s the way businesses prepare, adopt, integrate and govern AI. For leaders, this is both a wake-up call and a roadmap.

The best way network infrastructure managers and leaders can prepare for the next wave of AI is by focusing on the efficiency of their networks and the quality of their data.

Further, the Cisco Readiness Index Report shows that only 13% of companies globally are ready to leverage AI and AI-powered technologies to their full potential—a one-point decline from last year. As the report revealed, while relatively few companies report that they consider themselves prepared for AI, 98% said the urgency to use AI-powered technologies has risen in their organizations.

The evolution of AI

AI’s development phase has been interpreted in various ways, often remaining unclear about the relationship between machine learning, Gen AI and Artificial General Intelligence (AGI) and where one stops and the other starts. It’s often simply grouped into one big category: “AI”—which compounds the confusion.

AI includes everything from facial recognition on the low end to creativity, general wisdom and problem-solving beyond the capabilities of humans (called “artificial general intelligence (AGI)” by some, “superintelligent AI” by others) on the high end.

Where are we now in the evolution of AI? Most of the AI usage in today’s enterprise has been generative AI, which uses automation and machine learning to create new content, such as text, images and code from existing data. This stage of AI development began in the 2010s, when neural networks evolved into “deep learning”. Developments like computer vision, robotics and natural language processing, including the early large language models (LLMs), were introduced in this period.

Generative AI is beginning to mimic human creativity by creating new outputs that are similar to but not identical to the training data. The current wave of generative AI is promising to deliver autonomous agents capable of working independently, though the current technology has not yet matured to this point. But it’s close. The best expert systems are noticeably more advanced this year than last year. The best chatbots of 2025 can do infinitely more for us than the best chatbots of mid-2024. They’re expert systems, but they require clear guardrails and time to learn.

The state of AI in the enterprise network

Although network technology manufacturers have been embedding machine learning capabilities in their products over the past few years, their customers, enterprise network and infrastructure leaders, usually mean generative AI when they say “AI”. Typical use cases in this space have centered around routine, repeatable tasks:

• Reporting and data generation to support decision making
• Task automation
• Virtual assistants
• Optimization
• Issue detection
• Threat detection through pattern observation, scanning, patching etc.
  

Even for enterprises just wading in to gen AI, the potential benefits can be huge, though the risks associated with increased infrastructure workloads are real. But it’s the metrics that really matter: AI’s genuine usefulness to enterprise network managers will be in how it enhances the network’s efficiency and productivity and how those improvements can be measured.

Here’s what AI could mean for enterprise network managers

Automating complex tasks while embedding intelligent security features are some of the ways the integration of Gen AI and network management will revolutionize the field. Network managers will be able to focus on strategic initiatives, driving innovation and business alignment while driving digital transformation and enhancing user experience.

AI will play a pivotal role in ensuring peak network performance by automating key operational tasks. For example, intelligent network configuration can ensure optimal performance, while continuous monitoring can detect anomalies and inefficiencies in real time—eliminating the need for further escalations, even predicting and resolving potential failures before they impact operations. AI will help ensure smooth and efficient network operations by analyzing usage patterns and dynamically adjusting resources. Finally, to minimize waste and maximize productivity, AI can be used to facilitate smart resource allocation, including the distribution of computing power, bandwidth and storage efficiently and effectively.  

What’s next (and when)?  

AI is evolving in leaps and bounds. We're no longer talking about a distant possibility. The evolution from generative AI to AGI is accelerating, with incremental advancements building on one another. Some experts are predicting that AGI—still largely a theoretical concept—could emerge within the next decade. Others are saying 30 years or not at all, though we aren’t all using the same language. But until we reach AGI, a truly autonomous AI agent that would fully displace a human worker isn’t technologically possible.

And while there’s a lot of talk around this next stage of AI, very few have clear strategies in place for using it to its full capability. Let’s explore that.

The advent of AGI will bring about a monumental shift for enterprise network managers, transforming their roles and responsibilities in profound ways. Understanding AGI’s potential implications—including the challenges it presents—is crucial for future planning.

But until AGI becomes a reality, the best ways network managers can prepare for it are by focusing on optimizing network efficiency and data quality—because infrastructure must be optimized before AI can build on it effectively. Likewise, clean and well-structured data is needed to provide a strong foundation on which to build increasingly complex systems.

But are enterprise networks AI-ready?

As with any major transition, there are bound to be challenges in AI implementation and adoption. These include significant upfront investment, data privacy concerns, governance and ethical considerations, potential workforce changes and risks such as vendor lock-in. However, organizations that successfully tackle these challenges will gain significant competitive advantages.

Enterprise network and infrastructure leaders need to act now to meet the rising demands of generative AI and meet the requirements of AI workloads, including flexibility, scalability and smart environments. Many capabilities driven by AI, such as automation, predictive maintenance and adaptive security, are already crucial for managing these systems. Finding the right partner that can help businesses modernize, transform and improve their infrastructure for this AI-driven future is paramount to support seamless implementation and informed adoption.

In conclusion, AI may hold the potential to revolutionize enterprise network management, leading to unprecedented levels of automation, enhanced security and efficiency. However, it will also require network managers to navigate significant challenges. Network management in the age of AI will be about strategic leadership, human-AI collaboration and a focus on driving business value through intelligent and autonomous network infrastructure.