CivBench benchmarks LLMs playing Civilization V

Researchers introduced a controlled version of CivBench, having LLMs act as strategists in Civilization V across three fixed starts. GLM-5.3 outperformed Opus-5.5 and Qwen-3.8-27B held up surprisingly well; the benchmark tests long-horizon decisions whose effects appear 50-100+ turns later. The open-source Vox Deorum tool supports local models and human-vs-AI play.

• Agent harness affects cache efficiency and energy use2026-10-17
• Anthropic's Jack Clark moves AI self-improvement forecast to 20272026-10-15
• Engineer Says Anthropic's Scion Heavily Borrows from Mononoke2026-10-14
• Anthropic exec on safety-first AI strategy2026-10-14
• Open Models Debate: Weighing Safety Against Democratization2026-10-13