I Taught a Little and Learned a Lot at CMS Connect 25

I had the pleasure of attending CMS Connect 25 in Montreal with fellow Agility CMS employees Jina, Jessie, and Brendan. Now, here’s my attempt at deciphering my handwritten notes into something coherent.

AI is Everywhere. But What Does Real Adoption Look Like?

When I arrived at CMS Connect 25, I had one burning question: Where is CMS going?

Of course, I had my own idea heading in, which I presented during my talk: Skating to Where the Puck Will Be: CMS in the Age of AI.

By the end of day one, it was clear that AI was going to dominate most of our conversations about CMS.

Don’t get me wrong, the BASICS of CMS and headless are still important topics to most customers.  Structured content is a benefit that MOST organizations don’t benefit from enough. But AI is where it’s at in terms of the future of CMS.

AI was talked about everywhere. The big keynotes. Lightning talks. Hallway chats. Lunch conversations. Workshop sessions. AI was the dominant topic at CMS Connect 25 in a variety of ways:

• How can AI help with content generation?
• Can AI pull meaningful analytics?
• Is AI the panacea of personalization?
• Should AI be used for content governance?

There were even discussions on whether AI is capable of emotional storytelling.

Silent AI Adoption

It’s kind of crazy how AI has been adopted so fast, faster than other emerging technologies.

I think a big part of that adoption rate growth comes from the fact the many users are “silent” adopters. 

What does this mean? Silent adoption with AI happens when folks use the technology on the side, probably via ChatGPT.com, maybe even with a personal account. They’re writing emails, doing research, writing code, and the tooling is working better and better. Their trust in AI is growing, and users are starting to accept that AI will be a big part of all of their digital tools.

Now that AI is here, embedded into our daily lives, more questions popped up for me:

What do we do with it? How should it be integrated?

AI is on Every Roadmap

The appetite for AI in CMS is universal.

Customers are asking for it on RFPs as a “checkbox”, but they aren’t quite demanding it yet. We vendors, who are skating to where the puck is going to, are really looking for the best ways to integrate AI into our workflows.

Partners and developers are probably the biggest group with the most demand. These are technical folks who understand the value that AI can bring, and they're chasing the the AI edge more than anyone.

At CMS Connect 25, almost every session (mine included) had an AI angle. But the reality is that no single approach has emerged as the way to deliver AI to the enterprise. 

In previous innovation waves – headless CMS, SaaS, composability – best practices emerged quickly. This isn’t the case with AI. Everyone is still experimenting, from startups to large enterprises.

So far, nobody has cracked the code. But I feel like we’re getting close.

Experimentation is the New Normal

Right now, things feel different. Enterprises aren’t waiting to develop – or implement – fully baked features. Instead, they’re:

• Piloting AI features in real workflows.
• Running workshops to explore AI use cases.
• Trying beta features directly with various teams.

I’ve been with Agility CMS for 20 years, and I can seriously say I’ve never seen such broad, enterprise-wide experimentation. Let’s try it and see what happens is the prevailing, industry-wide attitude taken.

Excitement Meets Caution

Of course, there’s a flip side.

In her excellent presentation, Digital Governance in the Age of AI, Joyce Peralta of McGill University highlighted real concerns over trust, ethics, and control.

She also touched upon AI-washing – selling AI features in name only without adding real value.

Still, I came away convinced that most enterprise customers are ready to embrace AI if it fits within their workflow and demonstrably saves time or improves results.

It almost feels like a coin flip. Heads, and AI doesn’t help the way people work. Or tails, it does.

Workflow Fit Above All Else

If there’s one clear takeaway from vendors like Agility CMS, it’s this: AI adoption will accelerate when it fits seamlessly into existing processes.

Customers are not interested in a major overhaul to incorporate AI. Rather, they want AI to:

• Augment, not replace, existing workflows.
• Solve specific pain points they already feel (and not create new ones in the process).
• Be transparent, controllable, and trustworthy.

If an AI feature feels like it was bolted on or forces a team to change their process entirely, adoption will stall. But if it slips in seamlessly into what they already do, and makes things easier?

That’s gold, Jerry. Gold.

What Can the Community Do?

• Be a safe space to test AI. Nobody wants to create Skynet. Provide a testing ground where pilot projects can be low-risk and high-learning.
• Focus on workflows. Using AI to generate an image or alt-text is low-hanging fruit. Automate, enrich, and analyze in ways that matter day-to-day.
• Make AI transparent and controllable. Give users the confidence to trust AI in production and to modify if needed.
• Don’t believe the hype. Yes, this is a Public Enemy reference. Build what delivers measurable, repeatable outcomes and not just a cool feature that adds no value.

Final Thought and Some Thank Yous

To summarize everything I just wrote above:

AI is on everyone’s roadmap, but no one has perfected the perfect rollout playbook yet. For the first time, big enterprises are willing to experiment at scale. Vendors who focus on workflow fit and real user outcomes will win.

Finally, thank you to Janus Boye of Boye & Co. for once again organizing and hosting CMS Connect 25. Every community needs a champion, and the CMS community has a great one in Janus.

Thanks also to all attendees, speakers, and presenters for sharing their insights and knowledge. Together, we listen. Together, we learn. Together, we grow.