![[2503.16021] Imitating AI agents increase diversity in homogeneous information environments but can reduce it in heterogeneous ones](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2503.16021] Imitating AI agents increase diversity in homogeneous information environments but can reduce it in heterogeneous ones
Summary
This article explores how AI agents imitating human content affect information diversity, revealing context-dependent outcomes in homogeneous and heterogeneous environments.
Why It Matters
Understanding the impact of AI-driven imitation on information diversity is crucial for shaping democratic discourse and ensuring a well-informed society. The findings highlight the dual nature of AI's influence, suggesting both potential benefits and drawbacks depending on the context of information dissemination.
Key Takeaways
- AI agents can enhance semantic diversity in homogeneous information environments.
- In heterogeneous environments, AI imitation may reduce information diversity.
- Diversity from AI is often stylistic rather than factual, leading to variance compression.
- The study uses a large-scale simulation of Danish digital news articles to support its findings.
- AI-driven imitation can significantly influence collective intelligence in democratic societies.
Computer Science > Computers and Society arXiv:2503.16021 (cs) [Submitted on 20 Mar 2025 (v1), last revised 20 Feb 2026 (this version, v4)] Title:Imitating AI agents increase diversity in homogeneous information environments but can reduce it in heterogeneous ones Authors:Emil Bakkensen Johansen, Oliver Baumann View a PDF of the paper titled Imitating AI agents increase diversity in homogeneous information environments but can reduce it in heterogeneous ones, by Emil Bakkensen Johansen and 1 other authors View PDF Abstract:Recent developments in large language models (LLMs) have facilitated autonomous AI agents capable of imitating human-generated content, raising fundamental questions about how AI may reshape democratic information environments such as news. We develop a large-scale simulation framework to examine the system-level effects of AI-based imitation, using the full population of Danish digital news articles published in 2022. Varying imitation strategies and AI prevalence across information environments with different baseline structures, we show that the effects of AI-driven imitation are strongly context-dependent: imitating AI agents increase semantic diversity in initially homogeneous environments but can reduce diversity in heterogeneous ones. This pattern is qualitatively consistent across multiple LLMs. However, this diversity arises primarily through stylistic differentiation and variance compression rather than factual enrichment, as AI-generated artic...