Substitutability of conversational AI for traditional search on political issues

Establish whether large language model-based conversational AI systems function as practical substitutes for traditional internet search engines when users research common political issues.

Background

The paper notes ongoing debates about the reliability of LLMs, including hallucinations and potential political bias, alongside recent technical mitigations such as retrieval-augmented generation and fine-tuning for factuality. Within this context, the authors explicitly state that it remains unclear whether conversational AI can serve as functional substitutes for search engines for typical political information-seeking tasks.

References

Second, it is unclear whether conversational AI systems present functional substitutes for traditional search engines on common political issues.

— Conversational AI increases political knowledge as effectively as self-directed internet search  (2509.05219 - Luettgau et al., 5 Sep 2025) in Introduction, paragraph 3

Against that benchmark, the small AI surplus reads as residual selection, not an effect. The episode-level comparison remains the right default; the substitution question stays open only at the semantic level the trace cannot reach.

— Event-Time Confounding Under Bursty Human Dynamics  (2608.21294 - Iannelli et al., 21 Aug 2026) in Section 6, paragraph “The episode-level comparison, audited”