When someone sends me a thoughtful message, I think that is really sweet and kind. It means they have taken time for me. When I suspect AI helped prepare it, though, something changes psychologically, even when I can understand why they used it. I think there should be some indication of AI involvement in almost every type of message.
I hadn’t talked to my brother in two decades when I sent him a message out of nowhere. His first response read to me as clearly AI-assisted. That was my impression, rather than something he had confirmed, but the messaging felt disingenuous and prescribed. I could understand why he might use help to make sure the tone was correct, and it still felt flat on a personal level.
AI is changing work. Subscribe to The Human Part to explore what that asks of us as leaders.
What acknowledgement could have changed
I think it would have felt more personal if he had acknowledged the help. Something along the lines of, “I used AI to help me create a tone that I felt comfortable with.” Those are the words I imagine appreciating, not something he actually said.
That acknowledgement would have given me a different way to understand the reply. I could see someone trying to get the tone right in a difficult exchange. Without it, I was left with my impression of the wording and how impersonal it felt. Neither impression tells me everything about what he intended.
There is a reasonable argument for using the help. When the words matter, getting them wrong can matter too. I can understand wanting assistance with a message you don’t know how to begin. My discomfort with the result doesn’t mean the person had nothing sincere to say.
Transparency can be uncomfortable
Research complicates my preference. In two preregistered studies, Bolin Cao and Zhenming Li found that cues indicating AI involvement lowered perceived sincerity in apologies and expressions of gratitude, even when the message itself stayed the same. Participants were judging constructed student messages while imagining themselves as university instructors. These were brief scenarios, not evidence about what happens over time between family members.
The finding makes acknowledgement less comfortable to recommend. Telling someone AI helped could change their response to words they otherwise would have appreciated. It also leaves a question about the recipient: are we responding to what the person meant, or to what we think the tool’s involvement says about them?
I think there should be some indication of AI involvement in almost every type of message.
Tell me how the words came together
I still think acknowledgement is worth making. Transparency gives the recipient something real to work with, including the chance to understand why someone wanted help. It doesn’t require pretending the words will feel the same afterward, and I wouldn’t promise that a disclosure makes a difficult exchange easier.
For me, the personal part includes being open about how the message came together. I think that transparency is a huge win. If AI helped you prepare the words, I would rather you tell me than leave me wondering.
This article was prepared with AI assistance from my interview answers.
Dr. James Kelley
drjameskelley.com
Get The Crucible’s Gift on Amazon
Check your AI fluency
If this gave you something to think about, I’d love to have you back. Subscribe for more on AI, leadership, and the human part of work.
Sources
Cao and Li, The sequential replicant effect in interpersonal sincerity: evidence from AI-mediated apology and gratitude. Published online October 8, 2026. Message-cue experiments; perceived sincerity in constructed situations, rather than actual motives or long-term family relationships.


