In my first few months at Cisco, I’ve had the chance to dive into a lot of what’s happening behind the scenes – there’s been no shortage of innovation to take in. But one theme that has quietly stood out was something that’s less about what we can build with AI, and more about how we build it.
Cisco’s Responsible AI Framework isn’t a side project, it’s a central pillar of the business. And in today’s climate, that might be the most important move we’re making. Ultimately, it comes down to repatriation. Amazon, Google and Microsoft spent the last decade telling companies to move to the cloud. Now, some of those same workloads are coming back. But this isn’t a reversal, it’s a recalibration. Especially when it comes to AI.
Models are getting larger, data is more sensitive, and regulations are evolving faster than policies can keep up. Enterprises want the elasticity of cloud, but they also need the security, performance, and control of an infrastructure they trust. They’re rethinking not just how they run AI – but where.
What we’re seeing at Cisco is the rise of hybrid AI. But not just hybrid cloud – hybrid deployment, hybrid inference and hybrid control. The goal is to give teams the freedom to run models where they make the most sense – in the cloud when you need massive scale or on-premise when you need low latency, local compliance, or tight cost control.
Repatriating AI workloads isn’t about moving backwards. It’s about designing for intentionality – placing models closer to critical apps, complying with sovereignty requirements, or avoiding cloud sticker shock after every training cycle.
Cisco is building for this future across AI-native networks, workload mobility and secure data fabrics. And these aren’t buzzwords – they’re the connective tissue enabling AI to move to the edge, into private data centres, or anywhere performance and governance demand it.
For customers building their own models, protecting proprietary data, or navigating high-regulation environments, location is part of trust. Where your model lives, where your data flows, and where your decisions happen – all matter. That’s what makes Cisco’s Responsible AI Framework so relevant. It’s about more than principles, it’s about engineering AI that’s as thoughtful and flexible as it is powerful.
There’s a lot I’ve picked up in these first few months. I’ll leave it here for now and share more soon.
#CiscoAI #SecuredAI #AIDefense #FoundationAI #ResponsibleAI
