CI Conversations: What Improvement Practitioners Are Really Saying About AI

This week marked the first monthly ‘CI Conversations’ session from the Lean Competency System (LCS). An informal space for improvement practitioners to come together, share challenges, and learn from each other.

The topic for our first session was: AI in Continuous Improvement.

Below we have pulled six of the key themes that were discussed.

1. AI is already being used but often not where you expect

While AI is often positioned as a future capability, many practitioners are already experimenting with it in practical ways:

• • Process mapping support
    Using tools like AI-enabled whiteboards to generate high-level process maps from existing documentation
  • Turning documentation into insight
    Converting SOPs and guidance into structured process views as a starting point for improvement
  • Process design guidance
    Using AI to suggest best practice controls, risks, and process structures particularly in complex environments like finance
  • Governance and consistency checks
    Identifying mismatches between process maps, work instructions, and service documentation
  • Predictive applications
    In more advanced environments, AI is being used for forecasting and predictive maintenance

2. There’s still confusion: AI vs automation vs good data

One of the most consistent themes across the discussion was a lack of clarity:

“Some people are describing things as AI when it’s really just automation… or just better use of data.”

This matters because without clear definitions:

• Expectations become unrealistic
• Investments become misdirected
• And teams risk solving the wrong problems

As CI practitioners pointed out, many organisations are forgetting one of the fundamentals of CI:

You can’t have effective AI without good data and governance.

3. The biggest risk with AI is losing the human element of CI

Perhaps the most powerful discussion centred on a growing concern that if AI can generate a process map in 30 seconds, what happens to the collaboration and thinking that occurs mapping that in the traditional way through involving people?

Practitioners highlighted a real tension:

• CI has evolved to be collaborative, inclusive, and people-driven
• Many improvement tools (SIPOCs, fishbones, workshops) are valuable because of the conversations they create
• Automating these risks turning improvement back into something done to people, not with them

As one participant put it: “This misses the point… the value is in doing it together.”

We are all in agreement that efficiency gains should not come at the expense of engagement and ownership.

4. AI works best as an “augmenter,” not a replacement

Across multiple examples, a consistent pattern emerged:

• AI can suggest, but humans must validate
• AI can accelerate, but humans must decide
• AI can structure thinking, but humans must apply judgement

Even in areas like process mapping and design:

• Outputs still require sense-checking
• Context and nuance still matter
• Over-complexity can still be introduced without human intervention

The role of the practitioner is simply evolving, rather than at risk of disappearing. 

5. CI thinking is more relevant than ever

Interestingly, many participants reflected that CI principles are actually more important in an AI-enabled world:

• Start with customer value - Not everything should be automated
• Apply structured thinking - Governance, standards, and clarity are essential
• Focus on the problem, not the tool - AI is not the solution to all problems, but might be an enabler in helping solve them
• Build capability, not dependency
  Especially in learning environments, where over-reliance on AI can undermine skill development

One example raised the challenge in training environments:

Where is the line between someone demonstrating improvement capability and simply being good at prompting AI?

6. The opportunity: smarter, faster, more accessible improvement

Despite the concerns, there was clear optimism that if used thoughtfully:

• AI can help teams do more with less
• It can reduce manual effort and “donkey work”
• It can make CI tools more accessible to non-experts
• It can unlock new insights from existing data

Final Reflection: It’s not about AI vs CI, it’s about how they work together

What stood out most from this session wasn’t a single tool or use case, but the mindset that we adopt as practitioners.

AI is not replacing Continuous Improvement. If anything, it’s reinforcing its importance.

The organisations that succeed will be the ones that apply AI with the strongest CI thinking. 

Join the conversation

CI Conversations will run monthly, each session exploring a different theme shaped by the challenges practitioners are facing right now.

No slides. Just honest conversation.

If you’d like to join the next session, sign up here.