The moment an AI output stops being "something the tool suggested" and becomes your professional recommendation is easy to miss. This episode looks at that handover, when nobody double-checks, when a 'plausible' statistic almost ends up in a deck, and what it actually takes to know which outputs are reliable (rather than just hoping they are).
What this episode covers:
- Automation bias and why the risk arrives after practitioners feel comfortable with a tool
- The two types of task AI handles very differently: working with material you've provided versus generating from its own training data, and how the output looks identical either way
- What the ICO says about customer data in AI tools and what changes under the Data (Use and Access) Act 2025
- Three questions to ask before opening the tool, with examples from M&S and John Lewis Partnership on how responsible AI deployment looks at scale
Thanks for Listening!
Did anything resonate with you from today's conversation? I'd love to know whether the three questions landed for you and whether any of this shifted how you're thinking about what you put your name on.
Here's the link to the ICO Data (Use and Access) Act 2025.
Get the full show notes with all frameworks and resources mentioned at https://wheresyourcustomer.com/28
Never miss an episode – Subscribe to Where's Your Customer? on Apple Podcasts, Spotify, or wherever you get your podcasts
Found this valuable? Please leave a review, it helps other retail professionals discover these conversations and would honestly mean the world to me
Connect with me – Find me on LinkedIn or visit wheresyourcustomer.com for more customer service insights and resources.
Share this with someone who needs it – If someone in your network is wrestling with customer service challenges, please share this episode with them
