# Your AI Travel Agent Would Book You a Bullfight: An Agentic Benchmark for Implicit Animal Welfare in Frontier AI Models

arXiv: https://arxiv.org/abs/2606.18142 · HTML: https://mlsanity.com/paper/2606.18142/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Social simulation and multi-agent debate, one of 16 directions in LLM agents and tool use.
- Builds on 14 of its 27 references that are in the map.
- First posted on arXiv 16 Jun 2026.

## Placed in

- LLM agents and tool use / [Social simulation and multi-agent debate](https://mlsanity.com/agents/social-simulation-and-multi-agent-debate/)

## Builds on (most cited first, 14 of 14)

- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting](https://mlsanity.com/paper/2305.04388/) (2023)
- [Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training](https://mlsanity.com/paper/2401.05566/) (2024)
- [The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning](https://mlsanity.com/paper/2403.03218/) (2024)
- [Large Language Models Often Know When They Are Being Evaluated](https://mlsanity.com/paper/2505.23836/) (2025)
- [Probing and Steering Evaluation Awareness of Language Models](https://mlsanity.com/paper/2507.01786/) (2025)
- [Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment](https://mlsanity.com/paper/2601.10160/) (2026)
- [Why Are Web AI Agents More Vulnerable Than Standalone LLMs? A Security Analysis](https://mlsanity.com/paper/2502.20383/) (2025)
- [Evaluation Awareness Scales Predictably in Open-Weights Large Language Models](https://mlsanity.com/paper/2509.13333/) (2025)
- [Speciesist bias in AI -- How AI applications perpetuate discrimination and unfair outcomes against animals](https://mlsanity.com/paper/2202.10848/) (2022)
- [In-Context Environments Induce Evaluation-Awareness in Language Models](https://mlsanity.com/paper/2603.03824/) (2026)
- [Speciesism in AI: Evaluating Discrimination Against Animals in Large Language Models](https://mlsanity.com/paper/2508.11534/) (2025)
- [Untangling Input Language from Reasoning Language: A Diagnostic Framework for Cross-Lingual Moral Alignment in LLMs](https://mlsanity.com/paper/2601.10257/) (2026)
- [Do LLMs Hold Their Values? MANTA: A Multi-Turn Adversarial Benchmark for Animal Welfare Reasoning](https://mlsanity.com/paper/2605.16301/) (2026)

## Built on by (most cited first, 2 of 2)

- [Coercion and Deception in AI-to-AI Management: An Agentic Benchmark of Unprompted Escalation](https://mlsanity.com/paper/2607.15434/) (2026)
- [HarvestBench: Measuring Whether LLM Agents Will Pay to Avoid Killing Animals](https://mlsanity.com/paper/2609.04444/) (2026)

## Abstract (arXiv)

Previous research has evaluated animal welfare using question-and-answer benchmarks. This study investigates whether these evaluations also hold in agentic settings. The agents may showcase different behaviors compared to stand-alone large language models, as demonstrated in prior studies. This work introduces textitTAC (Travel Agent Compassion): the first agentic benchmark for assessing animal exploitation. TAC evaluates AI agentic behavior in travel booking scenarios across six animal categories, using thirteen hand-authored scenarios that vary by price, rating, and position, expanded via four augmentation variants into 52 prompts and run for three epochs, giving 156 scored observations per model. Nine frontier models across five model families were evaluated.. The results indicate that models tend to prefer harmful scenarios, performing below the random chance rate of 65% for selecting a neutral booking option, with Claude 4.8 achieving the highest performance at 64.7%. To address this issue, the persona of an ethical-brand identity was infused into the system prompt, resulting in welfare rates increasing from 32 to 80 percentage points, with a mean of 53 across all nine models. No evidence of evaluation awareness affecting the results was found, based on an Inspect Scout audit of 3,120 transcripts. These findings are directly relevant to the EU General-Purpose AI Code of Practice, which identifies non-human welfare as a systemic risk. TAC provides a practical method for measuring this risk.