Your AI Strategy Has a Human Problem with Christina Lewis
Most organisations are treating AI adoption as a technology rollout - and it's costing them. Christina Lewis, a specialist in emotional intelligence and human-centred change who has scaled companies like Deliveroo and Bolt, joins Rhiannon Stroud to unpack why so many AI transformations stall, and what to do instead.
Introduction
AI adoption has a people problem. Most organisations are pouring investment into tools and platforms while assuming the human side will sort itself out. Christina Lewis thinks that's exactly backwards - and the neuroscience of change resistance, the hidden cost of outsourcing our thinking to LLMs, and case studies ranging from Canva to Atlassian all point to the same conclusion: without emotional intelligence at the centre, AI transformations stall.
In this episode you'll hear:
- Why framing AI adoption as a tech implementation rather than a people transformation is so dangerous
- The neuroscience of why our brains resist change - and how to work with it instead of against it
- What "AI presenteeism" looks like, and why compliance isn't the same as adoption
- Why outsourcing too much thinking to AI risks eroding critical thought, creativity and originality
- How to "scaffold" AI use so people think first and use AI to challenge their ideas, not replace them
- What separates the organisations getting real ROI from AI from those quietly bleeding money on it
Christina brings sharp research, real case studies, and a clear framework for any leader serious about making AI adoption actually stick.
This isn't a tool. It's taking people through a grievance period, an identity shift - and that can't be done by just saying, 'Here's your login.'
Chapters
00:02 - Are we investing in people as much as tech?
00:36 - Christina's background scaling fast-growth tech companies
01:22 - What centering EQ in transformation looks like in practice
01:54 - Why treating AI adoption as tech, not people, is dangerous
02:53 - The neuroscience of why our brains resist change
04:27 - Can you rewire people's relationship with change?
05:33 - Why scale-ups can't skip the human side of AI rollout
06:13 - The Canva case study: unlimited tools, no direction
07:17 - AI presenteeism: when compliance looks like adoption
09:12 - The Wharton study: an 11% rise in boredom among AI users
10:03 - The $161bn problem: what fragmented AI use costs the Fortune 500
10:46 - Why AI belongs under the people function, not the CTO
11:15 - Giving everyone the keys to the AI Ferrari without teaching them to drive
12:56 - What managers must understand before tools are deployed
15:05 - Atlassian: only 24% of leaders use AI to improve collaboration
15:59 - Adoption vs adaptation: logins aren't impact
17:01 - "Keep what you kill": the reframe that helps people embrace AI
18:08 - Why efficiency gains get swallowed by more busywork
19:05 - Scaffolding: preserving critical thinking as AI takes over
22:50 - Does a high-EQ organisation scaffold better naturally?
23:14 - Why you can't prompt your way to genuine EQ
24:21 - Why culture is the real AI differentiator
25:01 - Three culture risks of getting AI wrong
27:00 - Why AI widens performance gaps and risks your best people
30:19 - Identity, grief and the SCARF model of change
32:44 - What winning organisations say to people about change
34:46 - Why enterprises face the hardest AI transformation problem
36:27 - Using AI to reimagine today, not automate yesterday
37:08 - The HelloFresh case study: reimagining core operations with AI
39:11 - Fast forward 3 years: what separates EQ-centred businesses
Key Take-aways
- Adoption isn't adaptation. Measuring logins, licences, or how many people have access to AI tools tells you almost nothing about whether your organisation is actually changing how it works.
- A team full of individual AI super-users doesn't automatically add up to a more productive team. Without shared context and direction, personal efficiency gains stay personal.
- Managers can't lead others through change until they understand their own relationship with it first. Rolling out tools before addressing this is why so many transformations stall at the middle layer.
- Outsourcing too much thinking to AI has a real cost. It risks eroding the critical thought, creativity and originality that make people - and their work - valuable in the first place.
- "Scaffolding" protects that thinking: form your own point of view first, then use AI to challenge and extend it, not replace it.
- Time saved by AI shouldn't be automatically refilled with more output. If efficiency gains just become more busywork, nothing has actually improved.
Need an Expert for your challenge?
We've sourced the best independent global talent so you don't have to. Every People and Culture specialism covered in one place by our trusted Experts. Search, book, brief and pay through one easy-to-use platform.
Explore Experts