As language models are deployed as autonomous agents that negotiate, cooperate, and compete on behalf of human principals, their strategic dispositions acquire direct economic consequences. Here we show, across 51,906 game-theoretic trials generating 826,990 strategic decisions from 25 large language models spanning seven developers and 38 canonical games, that models converge on competitive and coordination behaviour (coefficient of variation 0.06 for coordination, 0.11 for strategic depth) while diverging 48-fold on cooperation, from 1.5 per cent (GPT-5 Nano) to 71.5 per cent (Claude Opus 4.6). Provider identity is the dominant predictor of cooperative disposition, and this divergence is generationally unstable: OpenAI cooperation fell from 50.3 to 1.5 per cent across four model generations while Google cooperation rose from 8.3 to 56.8 per cent. Endgame analysis reveals that Anthropic frontier models sustain 57 per cent cooperation in the final round of finitely repeated games, where backward induction predicts zero, while the newest Google models cooperate throughout but universally defect when punishment becomes impossible. These strategic personalities are shaped by training pipelines, shift unpredictably across model versions, and cannot be inferred from capability benchmarks, yet they determine the cooperative outcomes of every economic interaction these models mediate. The complete dataset and an interactive explorer for the data are publicly available at https://felipemaffonso.github.io/strategic-personalities/.

翻译：随着语言模型作为自主智能体被部署，代表人类主体进行谈判、合作与竞争，其战略倾向便获得了直接的经济后果。在此，我们通过涵盖7家开发商的25个大语言模型在38个经典博弈中进行的51,906轮博弈实验，生成826,990个战略决策，表明模型在竞争与协调行为上趋于收敛（协调行为变异系数为0.06，战略深度为0.11），而在合作行为上存在48倍的差异，从1.5%（GPT-5 Nano）到71.5%（Claude Opus 4.6）不等。开发者身份是合作倾向的主要预测因子，且这种分歧在模型世代间不稳定：OpenAI模型在四代版本中合作率从50.3%降至1.5%，而Google模型则从8.3%升至56.8%。终局分析显示，Anthropic前沿模型在有限重复博弈的最后一轮中维持57%的合作率——此时后向归纳预测合作率为零——而最新的Google模型全程合作，但当惩罚不再可能时普遍背叛。这些战略特征由训练管线塑造，随模型版本不可预测地变化，且无法从能力基准中推断，但它们决定了这些模型中介的每一项经济互动的合作结果。完整数据集及交互式探索工具已在https://felipemaffonso.github.io/strategic-personalities/ 公开提供。