# Do AI reply suggestions change how people communicate?

Yes. Smart replies made chats faster and more positive and improved how partners rated each other, but suspected AI use was judged negatively.

Evidence: Moderate (Promising, test it with your users) · LLM fit: Low (Older AI research, check before applying) (https://aiuxevidence.com/how-we-rate)
Two randomized experiments with a pre-LLM smart-reply system.

## Design it like this

Offer suggestions that save time, and let users personalise them before sending.

- Expect suggestions to shift tone toward positive language. [from the study]
- Make suggestions easy to edit so messages stay personal. [our reading]
- Be careful with disclosure in personal contexts; suspected AI use carried a penalty. [from the study]
- Measure partner perception, not just reply speed. [from the study]

## What they found

### Hohenstein et al. (2021), experiment

Two randomized experiments (1,036 participants; Study 1: 438 MTurk workers) where pairs chatted with or without a commercially deployed smart-reply system. Using suggested replies made communication more efficient, with more positive language and more positive evaluations by partners. But people were evaluated more negatively when partners suspected them of using AI replies.

> “people are evaluated more negatively if they are suspected to be using algorithmic responses” (Abstract)

## Trade-offs and when it doesn't apply

- Pre-LLM smart replies in lab chats.
- Effects on long-term relationships not measured.
- Positive tone may not suit every message.

## What it looks like

- Do: Suggested replies that open in an edit field.
- Don't: One-tap canned replies for sensitive conversations.

## How to check it in your product

Survey message recipients on perceived warmth and authenticity with and without suggestions.

## Open questions

- Do LLM-written messages carry a bigger suspicion penalty?
- Should AI-assisted messages be disclosed?
- Does constant positive phrasing flatten communication?

## FAQ

### Do smart replies improve conversations?

Hohenstein et al. (1,036 participants) found more efficiency, more positive language and better partner evaluations.

### Is there a downside?

People were evaluated more negatively when suspected of using AI replies.

## Related questions

- [Should you label AI-generated content?](https://aiuxevidence.com/e/ai-labels)
- [Does writing with AI make everyone's content sound the same?](https://aiuxevidence.com/e/homogenization)
- [Can AI writing assistants change what users believe?](https://aiuxevidence.com/e/nudging-opinions)

## Sources

1. Jess Hohenstein, Dominic DiFranzo, Rene F. Kizilcec, et al. (2021). Artificial intelligence in communication impacts language and social relationships. arXiv:2102.05756v1 (2021); later in Scientific Reports. https://arxiv.org/abs/2102.05756

Canonical: https://aiuxevidence.com/e/smart-replies