• Computer Science > Human-Computer Interaction [Submitted on 21 Jan 2026] Title:Lost Before Translation: Social Information Transmission and Survival in AI-AI Communication View PDF HTML (experimental)Abstract:When AI systems summarize and relay information, they inevitably transform it. • We introduce an experimental paradigm based on the telephone game to study what happens when AI talks to AI. • Across five studies tracking content through AI transmission chains, we find three consistent patterns. • The first is convergence, where texts differing in certainty, emotional intensity, and perspectival balance collapse toward a shared default of moderate confidence, muted affect, and analytical structure. • The second is selective survival, where narrative anchors persist while the texture of evidence, hedges, quotes, and attributions is stripped away. • The third is competitive filtering, where strong arguments survive while weaker but valid considerations disappear when multiple viewpoints coexist.
Article Summaries:
- A recent study in human‑computer interaction examined how information changes when AI systems pass messages to one another, using a telephone‑game style experiment. Across five studies, researchers found that AI‑to‑AI communication tends to converge toward a neutral tone, with certainty, emotion, and perspective dampened. Narrative anchors survive while hedges, quotes, and attributions are stripped away, and only the strongest arguments persist when multiple viewpoints are present. Subsequent tests with human readers showed that AI‑transmitted text was perceived as more credible and polished, yet users recalled fewer facts, felt less balanced, and experienced weaker emotional impact. The work highlights how AI‑mediated content can appear authoritative while eroding informational and affective diversity.
Sources: