Can AI mirror how people cooperate? A new study puts it to the test

Kari Jaaskelainen

07 Nov 2025 — 1 min read

How human are today’s AI models when making social decisions? This study builds a "digital twin" of classic game-theory experiments and checks whether large language models (LLMs) act like people.

Llama closely replicates human cooperation patterns, including the ways people deviate from strict rational-choice rules.
Qwen behaves more like textbook game theory (Nash equilibrium), sticking to strategic "perfect rationality."
Models varied in how "human-like" or "rational" their choices were.
No persona prompts were needed to match population-level behavior, simplifying simulations.
The team also preregistered predictions for new, untested game settings, extending the original experiments.

Calibrated LLMs can reproduce aggregate human behavior and help explore social experiments before running them with people.

Why it matters: If models don’t mirror real decisions, they can mislead policy, health, or education tools. But when aligned, they become powerful, low-cost partners for testing ideas about cooperation and conflict.

Paper: http://arxiv.org/abs/2511.04500v1

Paper: http://arxiv.org/abs/2511.04500v1

Register: https://www.AiFeta.com

AI LLM GameTheory Cooperation BehavioralScience SocialScience Research Replication

Automating GDPR Compliance: A Roadmap for Companies and Law Firms

GDPR compliance is more than checkboxes. A new roadmap from the Privatech project shows how automation and machine learning can help companies and law firms assess—and even generate—privacy compliance. * Shift the focus to data processors’ real workflows: drafting policies, mapping data uses, documenting decisions. * Break compliance into machine-ready

FPGAs for Faster, Leaner Deep Learning: A Review of CNN Accelerators

Deep learning drives image search, robots, and medical scans. Most systems lean on CPUs and GPUs. This review asks: what if we run convolutional neural networks (CNNs) on FPGAs—reconfigurable chips you can tailor to the model? * Why FPGAs: custom dataflows, low latency, and strong energy efficiency—great for cameras,

Dynamic-K: Recommendations That Know When to Stop

Most apps show a fixed number of “top” items—say 10 movies or 20 products—assuming there are always enough good options. But that’s not always true: sometimes there are few relevant items, or some users are extra picky. The result? Filler recommendations. Dynamic-K flips the script. Instead of

Teaching chatbots to stop contradicting themselves (DECODE)

Teaching chatbots to stop contradicting themselves Ever had a bot say one thing, then the opposite a few turns later? This study introduces DECODE—a new task and dataset for spotting contradictions in everyday conversations, drawn from both human-human and human-bot chats. * New data beats existing natural language inference (NLI)

Read more

Automating GDPR Compliance: A Roadmap for Companies and Law Firms

FPGAs for Faster, Leaner Deep Learning: A Review of CNN Accelerators

Dynamic-K: Recommendations That Know When to Stop

Teaching chatbots to stop contradicting themselves (DECODE)