# Moral Mazes in the Era of LLMs

arXiv: https://arxiv.org/abs/2603.20231 · HTML: https://mlsanity.com/paper/2603.20231/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 26 references that are in the map.
- First posted on arXiv 6 Mar 2026.

## Builds on (most cited first, 7 of 7)

- [LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methods](https://mlsanity.com/paper/2412.05579/) (2024)
- [JudgeBench: A Benchmark for Evaluating LLM-based Judges](https://mlsanity.com/paper/2410.12784/) (2024)
- [Gmail Smart Compose: Real-Time Assisted Writing](https://mlsanity.com/paper/1906.00080/) (2019)
- [Subliminal Learning: Language models transmit behavioral traits via hidden signals in data](https://mlsanity.com/paper/2507.14805/) (2025)
- [Colors in Context: A Pragmatic Neural Model for Grounded Language Understanding](https://mlsanity.com/paper/1703.10186/) (2017)
- [Smart Reply: Automated Response Suggestion for Email](https://mlsanity.com/paper/1606.04870/) (2016)
- [Generative Text Steganography with Large Language Model](https://mlsanity.com/paper/2404.10229/) (2024)

## Built on by (most cited first, 1 of 1)

- [Need or Reputation? Region, National Income, and Academic Status in LLM Gatekeeping Decisions](https://mlsanity.com/paper/2608.05178/) (2026)

## Abstract (arXiv)

Navigating complex social situations is an integral part of corporate life, ranging from giving critical feedback without hurting morale to rejecting requests without alienating teammates. Although large language models (LLMs) are permeating the workplace, it is unclear how well they can navigate these norms. To investigate this question, we created HR Simulator, a game where users roleplay as an HR officer and write emails to tackle challenging workplace scenarios, evaluated with GPT-4o as a judge based on scenario-specific rubrics. We analyze over 600 human and LLM emails and find systematic differences in style: LLM emails are more formal and empathetic. Furthermore, humans underperform LLMs (e.g., 23.5% vs. 48-54% scenario pass rate), but human emails rewritten by LLMs can outperform both, which indicates a hybrid advantage. On the evaluation side, judges can exhibit differences in their email preferences: an analysis of 10 judge models reveals evidence for emergent tact, where weaker models prefer direct, blunt communication but stronger models prefer more subtle messages. Judges also agree with each other more as they scale, which hints at a convergence toward shared communicative norms that may differ from humans'. Overall, our results suggest LLMs could substantially reshape communication in the workplace if they are widely adopted in professional correspondence.