We came away from Cisco Live with a slightly different feeling than in previous years. Not excitement, exactly. More a sense of confirmation. Many of the things customers have been wrestling with quietly for a while were being spoken about openly on stage and in breakout rooms. Security and AI, especially, were no longer treated as parallel threads. They were spoken about as one problem space, whether that makes life easier or not.
Why this event mattered
Almost everything at the event circled back to two ideas: security and AI. Not separately, not politely taking turns, but overlapping in ways that felt deliberate. The sheer weight of attention given to them said a lot about where Cisco’s engineering effort and roadmap energy are going. And by extension, where large enterprise environments are likely to head over the next few years.
What struck us is that AI was no longer framed as something experimental. It was treated as a given. Networks are expected to carry AI workloads. Operations teams are expected to rely on AI-generated insight. Security teams are expected to protect systems that move faster and behave less predictably than before. There was very little “one day this might happen” language. It was more “this is already happening, so here’s how we cope”.
For customers, that changes the tone of the conversation. This was not about blue-sky ideas or distant horizons. It was about what needs attention now, across networks, operations, security, and skills, if organisations want to stay competitive over the next three to five years. Perhaps even sooner.
What stood out
Certain ideas kept resurfacing, sometimes in formal sessions, sometimes in quieter conversations around demos. AI-assisted operations came up repeatedly, with automation described less as a bonus and more as something teams will struggle without. That shift felt honest, if slightly uncomfortable.
There was also a clear push to extend these ideas beyond traditional IT. Industrial and operational technology environments were very much part of the discussion, with tighter links between Splunk, Catalyst Center, and Cyber Vision used as examples of how IT and OT management are being pulled closer together. It didn’t feel theoretical. It felt like a response to customers already trying, and sometimes failing, to manage these worlds separately.
The AI POD concept appeared often as well. Not as a marketing flourish, but as a practical building block for AI-ready infrastructure, pairing NVIDIA GPUs with Cisco Nexus switching. Secure access to AI services, stronger identity protection for both people and service accounts, and internal malware analysis services were also recurring topics.
What I appreciated, and this doesn’t always happen at large events, was the willingness to talk about friction. Skills gaps. Legacy systems. Governance worries. Operational risk. These weren’t brushed aside. If anything, they were acknowledged as the reason many customers move more slowly than vendors might like. That candour made the conversations more useful.
What Cisco is prioritising
A lot of Cisco’s messaging pointed towards becoming the underlying infrastructure layer for AI-driven organisations. That means AI-ready networks and silicon, scalable data centre designs, and hardware that pays attention to power use and cooling, not just raw performance. Unified fabric management and security architectures were spoken about as necessities rather than optional extras.
Secure Access featured heavily as a control point for how AI is used inside organisations. This includes visibility into desktop AI tools, SaaS-based AI services, and API-driven integrations. The emphasis was on knowing what is in use, applying policy consistently, and inspecting encrypted traffic where needed to spot threats. Data protection controls were discussed in practical terms, tied directly to internet and internal access policies.
Identity came through as a frontline concern. Not just multi-factor authentication for users, but proper protection for administrators and service accounts too. There was a noticeable shift away from treating non-human identities as an afterthought. They are now recognised as both powerful and risky.
Another area that drew attention was malware analysis delivered as a managed internal service. The idea is simple but important: letting teams analyse suspicious files without sending sensitive data to public platforms. It supports helpdesks, reduces hesitation when something looks wrong, and aligns better with privacy and regulatory needs.
Underlying all of this is what Cisco described as its AI Foundation. AI is meant to sit across security, collaboration, network operations, and automation, accessed through APIs rather than locked behind a single interface. That architectural choice feels sensible, though it does demand more thought from customers about integration and control.
What is real versus what is still taking shape
Some things are already very much deployable. AI assistants are live in Webex, security tooling, and networking platforms, offering summaries and plain-language troubleshooting. Secure Access with AI discovery is in use. Duo now covers administrators and service accounts more thoroughly. Malware sandboxing platforms are established. Webex integrations with ticketing and CRM systems are practical, not experimental. Meraki APIs are already being used for day-to-day automation.
Other areas are clearly still forming. Fully autonomous network operations are talked about with confidence, but they are not a switch you flip. Broad orchestration across every tool remains uneven. Deeply customised assistants that replace entire workflows exist more as ambition than standard practice. And while integration is improving, the idea of a single, perfectly unified view across everything remains, for now, incomplete.
That’s not a criticism. Most organisations will adopt these capabilities gradually. In fact, they probably should.
Practical implications for customers
When you strip away the stage lighting and keynote language, Cisco Live leaves customers with some very grounded questions. Do we actually know which AI services are being used, and by whom? Are service accounts protected as carefully as user accounts? Can staff report suspicious files without triggering compliance concerns? Are collaboration tools linked into business systems in ways that can be reviewed and audited? And are we governing AI use deliberately, rather than responding after something goes wrong?
These are not questions for a distant future. They are operational issues many teams are already facing, even if they don’t always label them as “AI problems”.
Where Damovo can help
Damovo’s role is to help turn these broad themes into something workable. That often starts with designing and deploying Cisco architectures that make sense in the real world, from campus networks to data centres that can support AI workloads without overreach.
There is also a strong focus on bringing operations and security closer together, so automation, visibility, and protection are part of everyday work, not bolted on later. Managed and professional services play a role here, helping teams operate with more confidence and less risk.
AI risk assessments aligned with the NIST AI Risk Management Framework help organisations understand where their exposure really sits, across people, processes, and technology. From there, remediation work can address security gaps, data protection concerns, model governance, and operational controls. Regulatory alignment follows naturally from that groundwork.
On a more practical level, this includes designing secure access and AI visibility approaches, strengthening identity environments for both users and service accounts, building compliant malware analysis services, and linking collaboration platforms into ticketing, CRM, and knowledge systems in ways that can be reviewed and trusted.
The focus is not on technology for its own sake. It is about making systems usable, governable, and secure.
Final takeaway
Cisco Live 2026 made one thing hard to ignore. Security, AI, identity, and automation are merging into a single way of operating. Organisations that treat AI as just another tool may find themselves underprepared. Those that weave it into secure access, identity controls, and operational workflows, with clear oversight, will be better placed.
The technology is already here. What will set organisations apart is not how quickly they adopt everything new, but how carefully they choose their foundations, and how well their security keeps pace with networks built for AI. For customers and prospects, now feels like the right moment to pause, take stock, and work with partners who can turn intention into outcomes that actually hold up.