“My CEO keeps coming and asking me how we are using AI in the SDLC!” – AI Enabled Delivery According to 50+ Tech Leaders

Over the last 10 days, I’ve attended four different events: Ctocraft London 2025 (the only one not explicitly focused on AI, yet AI dominated the corridor conversations), our own Expert Talks (kindly hosted by Disney), and two private gatherings for the Equal Experts Customer Network (one in Manchester and one in Amsterdam). In total, I spoke with more than 50 technology leaders, each at different stages of trying to adopt AI within their software delivery processes.

What We’re Hearing from Tech Leaders

Here are a few standout quotes, which reflect the spectrum of challenges and excitement around AI:

• “Everything is moving so fast I constantly feel as though I’m falling further behind.”
• “We are trying to get access for our developers but there is a 3-month governance process to get new tools approved.”
• “My CEO keeps coming and asking me how we are using AI in the SDLC!”
• “I’m using AI, and I know exactly what LLMs can and can’t do. It hasn’t really changed that much.”

I think there are three lessons that are worth taking from the opinions I’ve heard:

1. Everyone feels behind – You’re not alone in this, and it’s not too late to take action – but as a leader you do need to ensure that bureaucracy doesn’t hamper (safe) experimentation.
2. Beware false ceilings – Some teams and individuals say they’ve reached the limits of current AI tools, but often they’ve just reached the limits of their process for using these tools effectively.
3. Identify the value AI brings – It’s easy to start playing with models and tools, or just to be “seen doing something”, but as with any tool it’s only as good as the task you give it (avoid the “golden hammer”!).

A Quick Note on Staged Adoption

At a couple of these events, we shared a simple “staged adoption” model for AI Enabled Delivery—from small, bottom-up experimentation to full-blown organisational transformation. Lots of people found it useful for clarifying where they are and where they’d like to be. 

Click here to find out more about that approach: AI ED: From “Faster Horses” to “Motor Cars”….

Three Actions to Take Now

Based on what we’ve seen and heard, we suggest these three practical actions that can help you cut through the noise and make real progress:

1. Organise communities around learning
   
   • Instead of everyone rushing to experiment in the same narrow area (often coding), encourage a variety of experiments across the whole software delivery lifecycle.
   • Use the Theory of Constraints lens to identify your biggest bottlenecks—be it analysis, design, quality, or operability—and focus AI experiments there (different teams may have different bottlenecks).
   • Make sure teams share their learnings widely. Collaboration and open knowledge exchange are critical for AI to have an impact across your organisation.
2. Run a single accelerated initiative experiment
   
   • Pick a well-defined, “accelerated” project (possibly greenfield) where you can truly test AI’s potential end-to-end.
   • Work with people that have done this before, so you don’t end up going quickly, but in the wrong direction!
   • Use this initiative to uncover friction points in governance, security, and approvals—then remove those blockers before scaling AI to other projects.
   • Resist the urge to simply layer AI on top of an existing cross-functional team that’s already maxed out delivering business as usual.
3. Create your own AI Enabled Delivery playbook
   
   • Document the best practices and lessons learned as you go. Don’t wait until “we have it all figured out.”
   • A great example to reference is the DEFRA AI SDLC (published under the Crown Commercial Licence). Tim Howard’s blog post, GenAI & Software Development – A New Paradigm, introduced this to the world and explains its genesis.
   • Your internal playbook should be a living guide that demystifies AI for everyone from new hires to senior leadership, ensuring consistency and reducing duplication of effort.

Final Thoughts

It’s natural to feel like you’re lagging behind—especially when so much AI news is making headlines daily. But remember, most organisations are still finding their feet. By encouraging community-led experiments, running a dedicated accelerated initiative, and codifying your learnings in a playbook, you can create a strong foundation for AI Enabled Delivery.

If you’d like to chat more about any of these ideas—or share what’s worked (and hasn’t) in your own AI journey – please get in touch. We’re always keen to compare notes and help make sense of this fast-moving space. Please don’t hesitate to connect with me directly on LinkedIn to share ideas and AI journey’s there too.