How should we design AI customer journeys without fake empathy?

Design AI customer journeys around the level of human judgement each moment needs. Keep routine, low-emotion tasks transparent and transactional, rather than performing care. Route grief, distress, complex friction and consequential decisions to people, supported by clear behavioural expectations, factual review and accountable human ownership.

Match the interaction to the human need

AI customer journeys work best when the interaction style reflects the customer's real situation, not a default human-like persona. Routine tasks such as payment follow-ups or broadband troubleshooting may be better handled by a clearly non-human, efficient service that avoids small talk and false emotional cues. In those moments, a transactional AI can reduce embarrassment or friction because the customer is not being judged by another person.

Human support should remain the default for moments involving grief, distress, vulnerability or decisions with serious personal consequences. A customer cancelling a deceased partner's account, for example, needs human care rather than an automated decision tree. The boundary is a brand and service-design decision: identify where speed and consistency help, and where genuine human judgement is essential.

Sources: The Prompt: Are your brand guidelines ready to deliver behaviour?, The Prompt: Why brand matters more than ever in an AI world

Define behaviour, not only tone of voice

Responsible AI customer journey design needs behavioural expectations that guide what an AI can do under pressure, not only words, visual style or a tone-of-voice guide. AI systems increasingly create brand meaning through many small interactions, including late-night support conversations and automatically generated content. A friendly prompt alone cannot tell a system how to handle inconsistency, uncertainty or customer frustration.

Set clear boundaries for factual review, escalation, data use, accountability and the treatment of sensitive situations. Review whether AI output is accurate, useful and consistent with the organisation's values before it reaches customers. AI can make marketing and service more responsive, but a journey designed mainly as a cost-cutting mechanism can weaken trust when people need reassurance that someone is accountable for the experience.

Sources: The Prompt: Are your brand guidelines ready to deliver behaviour?, The Prompt: Why brand matters more than ever in an AI world

AI should make customer journeys clearer, not perform humanity

We believe AI customer journeys should be designed around intention, judgement and clear limits rather than an imitation of human empathy. Customers need useful, reliable interactions and a visible route to human support when the situation calls for it. We have applied that principle in AI-enabled creative and digital delivery, combining strategists, designers, writers and developers with AI workflows rather than treating generation as a substitute for judgement. For G42, we built a custom CMS and generative AI pipeline that enabled editors to develop image prompts and create more than 150 videos in days, while maintaining a structured approach to content quality. That combination of system design, creative direction and review is how AI can increase capacity without making experiences generic or falsely intimate.

Sources: We’re creating new experiences with AI – but we still need imagination, The Prompt: Why brand matters more than ever in an AI world, The Prompt: Are your brand guidelines ready to deliver behaviour?

FAQs

When should an AI customer journey hand a customer to a human?

AI customer journeys should hand customers to people when the interaction involves grief, distress, vulnerability or a consequential personal decision. A human route should be a deliberate service boundary, not a fallback hidden behind repeated automated prompts. Routine tasks with low emotional stakes can remain automated when the AI is transparent about its role.

Can a transactional AI experience be better than a human-like chatbot?

A clearly transactional AI experience can be better when customers value privacy, speed and freedom from judgement more than conversation. In payment follow-ups, a non-human approach can avoid the discomfort of being contacted by a person. The right interaction style depends on the context and the customer's emotional need, not on making every AI sound human.

Are brand guidelines enough for AI customer journeys?

Brand guidelines are useful, but they are not enough to govern AI customer journeys. Tone of voice, content strategy and visual style shape expression, while AI also needs guidance for decisions, inconsistency, escalation and behaviour under pressure. Customer-facing AI should have clear expectations for accuracy, accountability and human review.

How can AI personalisation avoid feeling artificial?

AI personalisation feels less artificial when it responds to a real customer need rather than simulating a relationship. Relevant recommendations and context-aware communication can be useful, but generic automation disguised as intimacy can damage trust. Design for clarity about what the AI is doing, what it knows and when a person is available.