Softwarized Networks in the Age of Generative Artificial Intelligence: Use Cases, Challenges, and Opportunities

Software-defined networks (SDNs) have fundamentally transformed the networking industry over the past decade, giving network operators unprecedented flexibility to customize network behavior and automate network operations without needing to rely on equipment vendor development cycles. At the same time, generative artificial intelligence (GenAI) has been taking the world by storm, enabling (among other things) big leaps in programmer productivity. SDNs involve a complex programming and significant implementation aspect that today, in many cases, still limits what the network operators can practically achieve. This makes GenAI seemingly an ideal complement for SDNs with the potential of taking it to the next level. But can it? Although the application of GenAI to SDNs indisputably holds considerable promise, many unique challenges are yet to be well understood and resolved to separate hype from reality.