“The death of CMS” or a new beginning?

For many organisations, from marketing teams to IT leadership, the CMS is the very hub of their digital presence. It is where content is created, structured, and published. At the same time, we are now seeing AI tools that can create websites, generate code, and publish content automatically in just seconds.

So it is not surprising that the question of whether we are heading toward “the death of CMS” is being raised. And those of us working as digital partners to our clients have started asking ourselves the same thing.

After working extensively with well-known CMS platforms such as Umbraco and Optimizely, and closely following AI developments, I have noticed that most discussions tend to come down to three key questions. Let’s take a closer look.

1. If we keep our CMS, how can we use AI to develop it faster?

One clear trend right now is that development processes are changing. Concepts like vibe coding, where developers collaborate with AI to quickly create code, components, and solutions, are becoming a reality in most development teams.

Because of this, the question is no longer a simple yes or no regarding whether to use a CMS. Instead, it becomes about how we can develop faster in relation to the CMS we already have.

At the same time, many CMS vendors are in a somewhat unusual position. Many platforms are currently pushing a clear strategy to move customers toward their SaaS versions. The challenge is that these are often more expensive and sometimes more limited than the PaaS solutions many organisations already use.

In some SaaS versions, for example, you need additional add-ons or API layers, often at a significant cost, to enable AI functionality in the way modern development requires. Meanwhile, similar capabilities can be achieved in open-source solutions for just a fraction of the cost.

This naturally leads many organisations to question why they should pay for a heavy and costly SaaS solution if they still need to build much of the functionality themselves.

2. Can you build an enterprise website using only AI?

Today, AI can generate code, design, and structure for a website in minutes. But when we move into more complex environments, other needs quickly emerge:

• Multilingual support

• Integrations with business systems

• Product data

• Personalisation

• Security

• Content governance

In these cases, the question of whether a CMS is needed becomes far more complex.

Over the years, we have moved from monolithic CMS platforms to more headless and composable architectures. Today, there are many lightweight SaaS CMS options on the market with simple licensing models and fast implementation, which can work well for simpler informational websites.

However, these solutions are rarely sufficient for enterprise environments where significant business logic needs to be built around the system. In those cases, PaaS-based CMS platforms still have a strong role to play, offering something many lightweight SaaS solutions lack: a robust ecosystem for integrations, structure, and scalability.

The same applies to marketing automation and data-driven customer communication. Many organisations today use marketing automation platforms where website content forms the basis for personalised journeys, campaigns, and automated communication.

As AI becomes more integrated into these processes, for example by generating content, optimising messaging, or triggering actions based on customer data, the need for a stable and structured content source increases. Here, the CMS can continue to play a central role as the hub where content is created, organised, and distributed across channels and automated workflows.

3. Will AI make the editor role obsolete?

This is perhaps the most sensitive question for many people working with websites today. If AI can create all content, is there still a need for editors?

We are moving toward a world where users do not always navigate to websites. Instead, they ask a question in an AI service and receive the answer directly in the interface. In this scenario, something fundamental changes: users get the answer without ever visiting your website.

This does not necessarily mean your content has become worse, but rather that the traffic ends up somewhere else.

Going forward, both SEO and GEO (Generative Engine Optimization) will become central to editorial work. Content must be structured and published in ways that make it understandable not only for search engines, but also for AI models.

This leads to a new question: do we even need a CMS to publish content in this kind of setup? Or can AI manage the entire publishing chain, from creating GEO-optimised content to publishing it?

The answer is still unclear. But one thing is certain: content still needs to be structured, organised, and maintained. That responsibility does not disappear just because the technology changes. And that is likely where the editor’s role will continue to exist.

“The death of CMS” or not?

Despite all the questions raised by AI, I do not believe we are heading toward a full-scale “death of CMS.” What we are seeing, however, is a transformation in how CMS platforms are used.

Going forward, we will likely see more of:

CMS platforms acting as content hubs rather than publishing tools

AI as an advanced co-creator within editorial interfaces

Automated workflows where content is created, improved, and distributed with AI

Stronger connections between CMS, data, marketing automation, and AI models, where content is not only published but activated in personalised customer journeys

CMS platforms are unlikely to disappear, but they will not look the same as they do today.

Just as websites once evolved from static, hardcoded pages to dynamic CMS-driven platforms, we are now entering a new phase where AI becomes an integrated part of the entire content lifecycle.

And in that world, the most important question may not be whether CMS survives, but which CMS platforms manage to adapt first.