CIONET News

CIONET Trailblazer: Why Cisco's Rob Lay Says AI Needs a Parent, Not Just a Plan

Written by Peter Rew | November 17, 2026 @ 2:42 PM

We sat down with Rob Lay, Solutions Engineering Director & CTO UK at Cisco, for the latest CIONET Trailblazer conversation. Rob talked candidly about his own journey from AI sceptic to daily user, why the real risk to your career isn't AI itself, and why most organisations are still solving the wrong problem when they start their AI journey.

The Convert: You've been open about starting out as an AI sceptic. What changed?

I'll hold my hands up, for a while I treated generative AI as an enhanced Google, and I think a lot of our customers went through the same phase. What changed it was seeing what it could do with my own workload. Five years ago, prepping for a customer visit meant going to our account team and waiting a couple of hours for a summary of what they'd bought and where their strategy was heading. Now I drop the same request into our internal AI tools and have it back in about thirty seconds. People love to tell me AI is going to take my job. I don't think that's quite right: AI won't take your job, but somebody who knows how to use it better than you may well do.

The Governance: At a recent CIONET roundtable, one attendee compared generative AI to a clever toddler, saying "give it a box of matches and it won't be long before my kitchen is on fire." Is that fair?

Left to its own devices, that's probably close. There's a real divide between organisations that know exactly how they want to use AI and those still, at best, using it to summarise their emails from a week off. What's missing from the comparison is the parent, who steps in and redirects that energy somewhere sensible. It's the same with AI: call it human in the loop, or just the business staying close enough to understand what AI is actually going to generate for them. Either way, it's the guidance and direction which is key.

The Cyber Threat: You've said attackers only need to be right once, while defenders need to be right every time. How has AI changed that maths?

The window between a vulnerability being discovered and an exploit being built used to be measured in weeks and months. We're now measuring it in hours, if not minutes, if not seconds, because that's how fast AI models can identify and build an exploit. The old version gave organisations time to patch within their normal change cycles. That's very hard to do in a world that now runs at machine speed.

The Business Case: Where do most organisations go wrong when they start their AI journey?

Those that Cisco call “Pace Setters” treat AI as a business challenge first, technology second. That suits Cisco too: it gives us a clear picture of what a customer is trying to achieve, and we build the infrastructure around that. Cisco moved on from selling customers a box some time ago. What we help deliver now is a business outcome, which makes us a partner rather than just a vendor.

The Advice: If you had to leave CIOs with one piece of advice, what would it be?

Take a strategic view of what you want to achieve with AI, because messing around at the edges will never get you there. It's great that I can summarise my emails from the week I was on holiday, but that's not a real business benefit. Figure out the use case your organisation is genuinely going to achieve something with, and drive towards that.

Watch the full conversation with Rob Lay on the CIONET Trailblazer podcast.