PrivacySIM: leaderboard
Large language models (LLMs) are increasingly used to simulate human behavior, but their ability to simulate individual privacy decisions is not well understood. In this paper, we address the problem of evaluating whether a core set of user persona attributes can drive LLMs to simulate individual-level privacy behavior. We introduce PrivacySIM, an evaluation suite that benchmarks LLM simulation
Metric: acc(E_u+A_u) (self-reported). Source: benchmarklist.com. Saturation forecast: Around July 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 40.39 | #80 |
| 2 | Gemini 3 Flash | 40.27 | #99 |
| 3 | GPT-5.4 (Medium) | 39.3 | #79 (GPT-5.4) |
| 4 | GPT-5.4 (Low) | 38.77 | #79 (GPT-5.4) |
| 5 | GPT-5.4 Mini | 37.99 | #212 |
| 6 | Qwen 3.5 122B A10B | 37.34 | #173 |
| 7 | Qwen 3.5 27B | 36.41 | #187 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=privacysim · How It Works · Data refreshed daily, snapshot 2026-10-20.