# The Cost of Synthetic Empathy **URL:** \ **Category:** Using AI\ **Created:** [December 22, 2025, 4:00pm UTC](https://forum.zurb.com/t/the-cost-of-synthetic-empathy/460 "2025-12-22T16:00:44Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![Bryan](https://yyz2.discourse-cdn.com/flex034/user_avatar/forum.zurb.com/bryan/32/448_2.png) [@Bryan](https://forum.zurb.com/u/Bryan)\ **Post date:** [December 22, 2025, 4:00pm UTC](https://forum.zurb.com/t/the-cost-of-synthetic-empathy/460/1 "2025-12-22T16:00:45Z")
Empathy without action erodes trust. We love this recent article by @Tim_Thijsse on what he calls synthetic empathy. AI has become very good at sounding caring through tone, language, and emotional recognition, but often lacks the authority to fix problems. Synthetic empathy is where customers feel understood but still walk away unresolved, which quietly erodes trust and loyalty even as metrics like CSAT look healthy. Check out his article: [The (C)X-Mas Spirit: Where AI Empathy ends and Human Compassion begins](https://medium.com/design-bootcamp/the-c-x-mas-spirit-where-ai-empathy-ends-and-human-compassion-begins-8e1db2ab5db9) AI can detect signals, route issues, and scale responsiveness. But in moments with emotional weight, ambiguity, moral trade-offs, or lasting consequences, humans must step in. To make this practical, Tim created a CX decision matrix that helps teams decide where AI should lead and where humans must take ownership. ![image](https://canada1.discourse-cdn.com/flex034/uploads/helio/original/1X/18c8e2f5dc8c90a851c19c3ae7ce9e67886eb87b.jpeg) His message is simple: empathy is recognition, but compassion means taking responsibility and acting. AI can scale recognition, yet high-stakes moments still require human judgment and follow-through. 💬 **Here’s my question..** at what point does empathetic language produced by AI become a liability instead of a strength? ---
**Author:** ![ben](https://yyz2.discourse-cdn.com/flex034/user_avatar/forum.zurb.com/ben/32/442_2.png) [@ben](https://forum.zurb.com/u/ben)\ **Post date:** [December 22, 2025, 4:02pm UTC](https://forum.zurb.com/t/the-cost-of-synthetic-empathy/460/2 "2025-12-22T16:02:55Z")
> [@Bryan](#): > > 💬 **Here’s my question..** at what point does empathetic language produced by AI become a liability instead of a strength? It’s a liability from the beginning. _You’re not having to empathize with the user._ That’s the biggest problem. ---
**Author:** ![Bryan](https://yyz2.discourse-cdn.com/flex034/user_avatar/forum.zurb.com/bryan/32/448_2.png) [@Bryan](https://forum.zurb.com/u/Bryan)\ **Post date:** [December 22, 2025, 4:10pm UTC](https://forum.zurb.com/t/the-cost-of-synthetic-empathy/460/3 "2025-12-22T16:10:10Z")
> [@ben](#): > > You’re not having to empathize with the user. Agreed. I think there are [different flavors of empathy](https://www.linkedin.com/posts/bryanzmijewski_uxresearch-productdesign-productdiscovery-activity-7132929744504819712-_oKN). → Cognitive empathy, knowing how a person feels and what they might be thinking. → Emotional empathy, physically feeling what a person feels, like their emotions are contagious. → Compassionate empathy, understanding a person’s predicament, feeling with them, and spontaneously helping. ![image](https://canada1.discourse-cdn.com/flex034/uploads/helio/original/1X/df866e6f852b89567ba376ab22ebb2bc95e6b28a.jpeg) ---
**Author:** ![tim\_thijsse](https://yyz2.discourse-cdn.com/flex034/user_avatar/forum.zurb.com/tim_thijsse/32/791_2.png) [@tim\_thijsse](https://forum.zurb.com/u/tim_thijsse)\ **Post date:** [December 22, 2025, 5:08pm UTC](https://forum.zurb.com/t/the-cost-of-synthetic-empathy/460/4 "2025-12-22T17:08:35Z")
Interesting thought. Do we need to teain bots to emphasize? Do users like that bots say ‘I feel for you?’. Maybe this should already be the moment we need to switch to the human service employee? ---
**Author:** ![ben](https://yyz2.discourse-cdn.com/flex034/user_avatar/forum.zurb.com/ben/32/442_2.png) [@ben](https://forum.zurb.com/u/ben)\ **Post date:** [December 22, 2025, 5:53pm UTC](https://forum.zurb.com/t/the-cost-of-synthetic-empathy/460/5 "2025-12-22T17:53:52Z")
I think, yes, users do like bots that have an understanding of the person using them. But rather people like bots that will do things implicitly in the way they like. Example: “I’ll produce this information in charts _because I understand you_”. The gap comes from the core drivers (i.e. CEOs and executives) being too detached from their users, faltering in the realm of advocacy. From what I can tell, this happens very frequently as organizations get larger. And, if we’re leveraging AI to do the empathizing for us (as executives delegate in a similar manner), we fall to the same problem: not empathizing nor understanding our users. Great callout either way @tim_thijsse ! ---
**Author:** ![ben](https://yyz2.discourse-cdn.com/flex034/user_avatar/forum.zurb.com/ben/32/442_2.png) [@ben](https://forum.zurb.com/u/ben)\ **Post date:** [December 22, 2025, 5:56pm UTC](https://forum.zurb.com/t/the-cost-of-synthetic-empathy/460/6 "2025-12-22T17:56:15Z")
Curious, do we need to empathize on all levels to be an effective advocate? (my hunch is yes) ---
**Author:** ![Bryan](https://yyz2.discourse-cdn.com/flex034/user_avatar/forum.zurb.com/bryan/32/448_2.png) [@Bryan](https://forum.zurb.com/u/Bryan)\ **Post date:** [December 22, 2025, 7:12pm UTC](https://forum.zurb.com/t/the-cost-of-synthetic-empathy/460/7 "2025-12-22T19:12:49Z")
Going back to the different types of empathy, even among humans, most empathy in a business context is limited to cognitive empathy, which makes sense that companies believe AI can solve this challenge. Perhaps there’s another level of cognitive empathy that people are willing to accept with a bot. What signals do you think @tim_thijsse should trigger a human handoff?