Truace
Subscribe
68/100

Moderate evidence

Cite
More
Export CSVExport JSONEmailLinkedInX

Chatbot ads that speak as "I" hurt purchase intent, two experiments find

Study title: Agency cues in personalized chatbot advertising: the role of grammatical subject framing, privacy concern, and social role

Evidence type
Controlled or validated comparison
Sample
n = 698
Sources
1 sourcePeer-reviewed
Published
2 min read
Truvace record
TRV-2026-2513cite or check it

Study title: Agency cues in personalized chatbot advertising: the role of grammatical subject framing, privacy concern, and social role

Two online experiments tested how chatbot advertising works when it is personalized and when the sentence subject is the AI ("I") or the consumer ("you"). In Study 1, with 396 participants, highly personalized ads that used AI-subject framing increased privacy concern, which in turn lowered brand attitude and purchase intention.

Read the study abstract

This research examines how personalization and grammatical-subject framing in chatbot advertising shape privacy concern and advertising effectiveness. Across two experiments, we manipulated personalization (high vs. low) and grammatical subject (AI “I” vs. consumer “you”). Study 1 (N = 396) showed that, under high personalization, AI-subject framing increased privacy concern, which in turn reduced brand attitude and purchase intention. Study 2 (N = 302) replicated this moderated mediation pattern in a different…

The problem

In an experiment with 396 participants, highly personalized chatbot ads worded with the AI as grammatical subject ("I") raised privacy concern, which lowered brand attitude and purchase intention.

In brief

Study 2, with 302 participants, repeated the pattern in a different product category. It also tested the chatbot's social role, and privacy concern did more damage to advertising outcomes when the chatbot acted as a friend than when it acted as a servant.

The findings suggest that a small wording choice, who the sentence is about, can change how consumers respond to personalized conversational AI. Both studies were experiments, so they show effects under controlled conditions rather than measured sales.

Main points

  1. Study 1 had 396 participants and varied personalization (high vs. low) and grammatical subject (AI "I" vs. consumer "you").
  2. Under high personalization, AI-subject framing increased privacy concern.
  3. Higher privacy concern reduced both brand attitude and purchase intention.
  4. Study 2 had 302 participants and replicated the pattern in a different product category.
  5. Privacy concern had stronger negative effects on advertising outcomes when the chatbot played a friend rather than a servant.
  6. The authors describe grammatical-subject framing as a sentence-level agency cue.

The full story

The research was published in the International Journal of Advertising. It manipulated personalization at two levels and grammatical subject at two levels, comparing chatbot messages in which the AI spoke as "I" with messages addressed to the consumer as "you".

In Study 1 (N = 396), the pathway ran from AI-subject framing to higher privacy concern, then to weaker brand attitude and lower purchase intention. This held under high personalization.

Study 2 (N = 302) used a different product category and added chatbot social role as a boundary condition. Casting the chatbot as a friend strengthened the negative effect of privacy concern on advertising outcomes compared with casting it as a servant.

Sources

1 on file
  1. International Journal of Advertisingdoi:10.1080/02650487.2026.2749451
    Peer-reviewed2026-10-29

Talk it over

Four questions to start onWho carries this cost, and what would have to change to prevent it?

Start with a question

Chatbot ads that speak as "I" hurt purchase intent, two experiments findhttps://truvace.com/item/agency-cues-in-personalized-chatbot-advertising-the-role-of-grammatical-subject-

  1. Who carries this cost, and what would have to change to prevent it?

    The problem In an experiment with 396 participants, highly personalized chatbot ads worded with the AI as grammatical subject ("I") raised privacy concern, which lowered brand attitude and purchase intention.

    Answer in the thread

  2. How do we know? What can this kind of evidence show, and what can’t it?

    The evidence This rests on a controlled comparison with a sample of 698, from International Journal of Advertising.

    Answer in the thread

  3. What does it leave open? What else would you want to know before acting on it?

    The record’s own caveat The findings come from two experiments with 396 and 302 participants, covering two product categories, so they show effects under controlled conditions only.

    Answer in the thread

  4. Have you seen AI used this way where you study, work or live? Did it go the same way?

    Answer in the thread

The questions are the same on every record. The details under them are taken from this record as written.

What readers say

More from Business