How Utilitarian Are OpenAI's Models Really? Replicating and Reinterpreting Pfeffer, Krügel, and Uhl (2025)
Fuente:
arXiv
Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917359305883648 |
|---|---|
| author | Himmelreich, Johannes |
| author_facet | Himmelreich, Johannes |
| contents | Pfeffer, Krügel, and Uhl (2025) report that OpenAI's reasoning model o1-mini produces more utilitarian responses to the trolley problem and footbridge dilemma than the non-reasoning model GPT-4o. I replicate their study with four current OpenAI models and extend it with prompt variant testing. The trolley finding does not survive: GPT-4o's low utilitarian rate doesn't reflect a deontological commitment but safety refusals triggered by the prompt's advisory framing. When framed as "Is it morally permissible...?" instead of "Should I...?", GPT-4o gives 99% utilitarian responses. All models converge on utilitarian answers when prompt confounds are removed. The footbridge finding survives with blemishes. Reasoning models tend to give more utilitarian responses than non-reasoning models across prompt variations. But often they refuse to answer the dilemma or, when they answer, give a non-utilitarian rather than a utilitarian answer. These results demonstrate that single-prompt evaluations of LLM moral reasoning are unreliable: multi-prompt robustness testing should be standard practice for any empirical claim about LLM behavior. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_22730 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | How Utilitarian Are OpenAI's Models Really? Replicating and Reinterpreting Pfeffer, Krügel, and Uhl (2025) Himmelreich, Johannes Computation and Language Computers and Society I.2.7; J.4; K.4.2 Pfeffer, Krügel, and Uhl (2025) report that OpenAI's reasoning model o1-mini produces more utilitarian responses to the trolley problem and footbridge dilemma than the non-reasoning model GPT-4o. I replicate their study with four current OpenAI models and extend it with prompt variant testing. The trolley finding does not survive: GPT-4o's low utilitarian rate doesn't reflect a deontological commitment but safety refusals triggered by the prompt's advisory framing. When framed as "Is it morally permissible...?" instead of "Should I...?", GPT-4o gives 99% utilitarian responses. All models converge on utilitarian answers when prompt confounds are removed. The footbridge finding survives with blemishes. Reasoning models tend to give more utilitarian responses than non-reasoning models across prompt variations. But often they refuse to answer the dilemma or, when they answer, give a non-utilitarian rather than a utilitarian answer. These results demonstrate that single-prompt evaluations of LLM moral reasoning are unreliable: multi-prompt robustness testing should be standard practice for any empirical claim about LLM behavior. |
| title | How Utilitarian Are OpenAI's Models Really? Replicating and Reinterpreting Pfeffer, Krügel, and Uhl (2025) |
| topic | Computation and Language Computers and Society I.2.7; J.4; K.4.2 |
| url | https://arxiv.org/abs/2603.22730 |