As digital technologies increasingly shape children's early learning experiences, their role in supporting foundational literacy practices, such as shared reading, warrants investigation. The current article examined whether conversational agents (CAs) powered by artificial intelligence can enhance parent-child conversations that extend beyond the text (i.e., extratextual talk) and foster early literacy development. Forty parents and their 4- to 5-year-old children were randomly assigned to either a CA-assisted intervention group or a traditional shared reading (active control) group. Over 1 month, families in both groups read together at home at comparable frequencies. Results showed that CA-assisted reading was associated with increases in two key measures of extratextual talk: parent and child decontextualized language and conversational turn taking. Notably, some of these gains transferred to traditional, non-CA-supported reading contexts. Despite these interactional benefits, no differences were observed between groups on measures of children's early literacy skills at the conclusion of the intervention. These findings suggest that CAs can serve as effective tools for scaffolding parent-child conversation during shared reading. However, further research is needed to optimize CA designs to support long-term language and literacy outcomes. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
Leech et al. (Mon,) studied this question.