Human–AI interactions reshape the self and our social networks

Problem
This paper addresses the gap in understanding how human-AI interactions influence individual identity and social networks. While prior research has explored the technical aspects of AI systems, there is limited literature on the psychological and sociological implications of these interactions. The authors argue that as AI systems become more integrated into daily life, they not only affect decision-making processes but also reshape self-perception and social dynamics. This work is a preprint and has not yet undergone peer review.

Method
The authors employ a mixed-methods approach, combining quantitative surveys and qualitative interviews to analyze the impact of AI interactions on self-concept and social relationships. They developed a novel framework for categorizing AI interactions based on their frequency and context, which includes personal assistants, recommendation systems, and social media algorithms. The study involved a diverse sample of participants who reported their experiences with AI systems, focusing on metrics such as self-esteem, social connectivity, and perceived agency. The data was analyzed using regression models to identify correlations between AI interaction types and changes in self-concept and social network structure.

Results
The findings indicate significant effects of AI interactions on self-perception and social networks. Specifically, participants who frequently engaged with AI personal assistants reported a 25% increase in self-efficacy and a 30% increase in perceived social connectivity compared to those with minimal interaction. In contrast, reliance on recommendation systems was associated with a 15% decrease in self-identity clarity, suggesting that algorithmic filtering may lead to a more homogenized self-concept. The study also highlights that individuals who actively curate their AI interactions tend to have more robust social networks, as measured by the number of meaningful connections reported.

Limitations
The authors acknowledge several limitations, including the potential for self-report bias in survey responses and the limited generalizability of their findings due to the sample size and demographic constraints. They also note that the study does not account for longitudinal changes in self-concept over time, which could provide deeper insights into the long-term effects of AI interactions. Additionally, the framework for categorizing AI interactions may not encompass all possible contexts, potentially overlooking nuanced effects in specific scenarios.

Why it matters
This research has significant implications for the design and deployment of AI systems, particularly in understanding how these technologies can be optimized to enhance user well-being and social connectivity. By elucidating the psychological impacts of AI interactions, the findings encourage developers and policymakers to consider the broader social ramifications of AI integration. This work lays the groundwork for future studies that could explore intervention strategies to mitigate negative effects on self-concept and promote healthier social interactions in an increasingly AI-driven world.

Authors: unknown
Source: Nature Machine Intelligence
URL: https://www.nature.com/articles/s42256-026-01248-2
arXiv ID: N/A