Score: 1

AlignDP: Hybrid Differential Privacy with Rarity-Aware Protection for LLMs

Published: December 19, 2025 | arXiv ID: 2512.17251v1

By: Madhava Gaikwad

BigTech Affiliations: Microsoft

Potential Business Impact:

Protects smart computer brains from being copied.

Business Areas:
A/B Testing Data and Analytics

Large language models are exposed to risks of extraction, distillation, and unauthorized fine-tuning. Existing defenses use watermarking or monitoring, but these act after leakage. We design AlignDP, a hybrid privacy lock that blocks knowledge transfer at the data interface. The key idea is to separate rare and non-rare fields. Rare fields are shielded by PAC indistinguishability, giving effective zero-epsilon local DP. Non-rare fields are privatized with RAPPOR, giving unbiased frequency estimates under local DP. A global aggregator enforces composition and budget. This two-tier design hides rare events and adds controlled noise to frequent events. We prove limits of PAC extension to global aggregation, give bounds for RAPPOR estimates, and analyze utility trade-off. A toy simulation confirms feasibility: rare categories remain hidden, frequent categories are recovered with small error.

Country of Origin
🇺🇸 United States

Page Count
8 pages

Category
Computer Science:
Cryptography and Security