F5 CEO On ‘The Most Exciting Time’ In Two Decades As AI Accelerates App Delivery, Security: Exclusive
F5 President and CEO François Locoh‑Donou tells CRN how AI is driving a once‑in‑a‑generation shift in application delivery and security and how F5’s updated ADSP, expanded API security, and AI model protection capabilities are creating new growth opportunities for partners.
F5 President and CEO François Locoh‑Donou is positioning the AI era as a once‑in‑a‑generation inflection point for application delivery and security, with partners at the center of that opportunity.
As enterprises run more and more AI workloads across on‑premise, cloud, and edge environments, the complexity of delivering and securing apps and APIs is only accelerating. F5’s updated Application Delivery and Security Platform (ADSP) is designed to address that challenge by consolidating delivery and security into a single platform that spans hybrid and multi‑cloud environments. With customers now operating applications across an average of 19 environments, Locoh‑Donou said that the time for consolidation is now.
Locoh‑Donou spoke about how AI is reshaping the threat landscape and the urgency around API security and AI model protection. He pointed to F5’s 2025 acquisition of CalypsoAI as a key differentiator, which has granted F5 offensive and defensive AI security capabilities, alongside new on‑prem API security offerings that address gaps left by cloud‑only solutions. And looking ahead to 2026, Locoh‑Donou said partners should use refresh cycles as expansion opportunities to attach valuable offerings and new managed services, such as API security and AI data delivery.
In an exclusive interview from AppWorld 2026 with CRN, here’s what Locoh‑Donou had to say.
What is the big headline you want partners to take from AppWorld 2026?
I want [partners] to know that this is the most exciting time to be in the world of application delivery and security in probably two decades. Why do I say that? It’s because the trend of delivering and securing apps and APIs across hybrid and multi cloud environments is accelerating, and it requires more delivery and security, and it’s accelerating in large part because of AI. AI is a massive shift. A couple years ago, the thought was that most AI would be consumed in public clouds, in the clouds of the hyperscalers or the clouds of the model providers, and it turns out that enterprises don’t want to do that. For the most part, enterprises are actually managing AI workloads themselves, and whether they do that on-prem or in the cloud, those workloads need to be delivered and secured, and that is going to be the work of our partners working with F5.
The updated ADSP consolidates delivery and security capabilities into one place for enterprises. What’s the reason that consolidation is so important?
This uber trend of hybrid multi-cloud, customers have to be in multiple infrastructure environments to run their applications. In fact, we just did some research recently and found that our customers, on average, operate their applications in 19 different environments — data centers, co-location facilities and different cloud regions. And 86 percent of enterprises run their apps at the edge, in the cloud, and on-prem. So, when you start from that and say: OK, because of flexibility reasons, because of cost reasons, because of control reasons, customers are deployed in these hybrid, multi cloud environments, which is great for flexibility, but it brings enormous complexity with it. In a lot of cases, they have had to deploy these point solutions, oftentimes with different tools that are doing the same thing because it’s in a different location. And all of these tools have created massive complexity, and so the big appeal of the F5 ADSP is consolidation because we are converging all delivery and security capabilities into a single platform. And then on top of that, we’re making that same platform available across all these locations. For an operator, this consolidation brings a lot of simplification. And the reason why we’re seeing this traction is because customers want to operate in these hybrid, multi-cloud environments, but they don’t want the complexity that comes with it, and a platform like F5 really has the potential to significantly reduce that complexity.
The reason all of this stuff is accelerating is because hybrid multi-cloud itself as a trend, is being accelerated by AI, because AI, it turns out, is very hybrid. Customers have their models [and] their data, and typically, they want their data to be on-prem. They may have their models in the cloud. They may do some inference workloads at the edge or in the cloud, or sometimes it’s on-prem. And so, AI workloads themselves are hyper distributed, and they’re quite dynamic. Agentic AI is going to be very distributed across multiple locations. So, AI is accelerating the strength for hybrid multi-cloud, adding more fuel to the fire, if you will. It’s exacerbating the need for a platform. Then, there’s a third trend, which is digital sovereignty, and it’s particularly strong in Europe at the moment where there’s been regulation that dictates to a lot of businesses and government agencies how much resilience they must have in their architecture, how fast they have to recover from a disruption, where the data can live [and] which cloud infrastructure provider they can depend on, and those regulations are also creating more demand, and in turn, accelerating the need for a platform like ADSP.
---
How is AI changing the threat landscape for applications and APIs?
There are two places where [AI] is changing the threat landscape for customers. One is, [customers and partners] are taking API security a lot more seriously. Customers realize that anytime an AI application calls an AI model, or anytime an AI agent is going to call a tool, it’s an API that’s doing the work, so knowing where all their APIs are, knowing whether their APIs are configured properly and the data flowing in and out of these APIs are right, all of these issues around securing APIs have become really more important for customers in the last couple of years. The other big question is whether the AI models become themselves an attack surface. We’re seeing enterprises now, on average, have seven AI models in production. We thought they would standardize on one model, but it turns out they are going for seven models because different models do different things and do things better, but every one of these models has a lot of vulnerabilities, and those vulnerabilities need to be identified, and then they need to be remediated so you can protect the models. Securing AI is why we made the acquisition of Calypso AI in[September] and it’s given F5 both offensive and defensive capabilities for AI. Offensive, meaning the ability to identify vulnerabilities in models, to do penetration testing — it’s called red teaming — against models, and then be able to provide the right guardrails against these vulnerabilities. That is a place where we are starting to see quite a bit of demand.
---
What is the biggest security problem that F5 is intent on solving this year?
It’s securing applications and APIs across hybrid, multi-cloud environments, which is actually easier said than done. Let me give you a simple example of that. Essentially, the only solutions to secure and discover APIs today are cloud-based solutions from [vendors] like Akamai and maybe Cloudflare or some cloud players, but a lot of customers, most of their enterprise traffic is on-premise. In fact, their most sensitive enterprise traffic is on-premise, and a lot of them don’t know how many API’s they have even on-premise, and what these APIs are doing, and there hasn’t been a solution to deal with API discovery and API security on-premise. F5 is announcing a solution for securing APIs on-prem. That’s one example. AI security is the same thing. The AI security solutions today from cloud providers, they work in their clouds, but if you’re going to deploy seven models … you may have two or three of these models on prem [or] in some other cloud. To have a solution that can find the vulnerabilities of all these models and secure all these models? That’s not easy, and that’s what F5 does. We have a solution that works across any cloud for any model.
The biggest problem we intend to solve this year is securing all apps, all APIs, all AI models across any infrastructure environment, or across hybrid, multi-cloud environments. That is the mission of F5 and one that I think we are essentially the only ones capable of taking on right now, because everybody else is either 100 percent on-premise or 100 percent in the cloud.
Talk about the vendors that F5 is displacing today.
We’re still doing a lot of Citrix displacements. That has been quite a lot of fun, I must say, and it continues unabated. Sometimes we’re displacing smaller competitors that are part of these point security solutions — companies like Radware or A10 [Networks] — and there are places where we are displacing a SaaS competitor, like an Imperva or even Akamai, where the customer was using them in the cloud, and then they chose to use F5. There’s also a number of customers that are now moving away from VMware. VMware has a software ADC from Avi Networks, and we are also seeing some displacement of that. In the case of Citrix and VMware, both, it’s self-inflicted by their own behavior. It’s more customer dissatisfaction that has led to opportunities for F5.
Where should partners be focusing on with F5 in 2026 and what results would you like to see by this time next year?
I think a lot of partners of F5 are having natural conversations with customers at this moment because a lot of customers are refreshing their F5 install bases. The most natural motion for them to have is to attach new footprint to these refresh opportunities. The new footprint could be API security. So, selling F5 Distributed Cloud with API security. It could be adding an AI data delivery use case to their big IP refresh. A lot of customers actually now need our ADCs in front of data stores, which they did not need before, but they’re doing that for AI data pipelines. So, finding a way to expand beyond just a simple refresh, to attach an AI data delivery use case and taking every opportunity, every refresh, to start an AI conversation around AI security. The last one would be also taking the opportunity for refresh to expand into our platform play and bringing in some capabilities from the ADSP that a customer doesn’t have today. It could be a cloud service, or it could be better visibility and observability from a cloud console that starts to give the customer a single pane of glass to look at all of their security policies and all of their application performance across their hybrid, multi-cloud environments. I think the most natural conversation for our partners is when they’re engaging in these refresh opportunities and expanding that into WAF as a service, distributed cloud, API security, AI delivery or any platform services.
I think by this time next year, I would like to see hundreds of ADSP customers, and we measure those by how many services they’ve adopted. We have criteria to define what is a platform customer [and] they typically have adopted multiple services, in delivery and in security, to spend a certain minimum amount. The biggest measure for me would be to see the number of platform customers grow substantially over the next 12 months.