The Distribution Hypothesis: Empirical Evidence That Fragment Allocation, Not Fragmentation, Determines AI Decision Protection

The Distribution Hypothesis: Empirical Evidence That Fragment Allocation, Not Fragmentation, Determines AI Decision Protection

This paper presents FLEET (Federated Leakage Evaluation & Evidence Test), the first systematic benchmark measuring reconstruction attack success against semantic fragmentation across heterogeneous large language models. Over 4,350 attack trials spanning 50 queries across 5 high-consequence verticals (finance, healthcare, legal, government, insurance), 8 attacker configurations from 6 AI providers, and 3 statistical runs, we establish a critical distinction between fragmentation and distribution as defense mechanisms.

Key findings:

• Level 1 (Direct Attack): Fragmentation alone provides zero protection — 0 of 1,200 trials achieve Decision Non-Reconstructability (D_A < 0.3; avg similarity 0.922). This result holds across fragment counts k ∈ {2, 3, 4}, eliminating fragment granularity as a tunable defense parameter.
• Level 2 (Collaborative Attack): Controlled fragment distribution reduces reconstruction success by two orders of magnitude — 81.3% ± 3.1% of trials achieve D_A (avg similarity 0.224).
• Residual Scope: 3 queries (6%) fail D_A consistently across all runs; an additional 7 queries (14%) exhibit stochastic failures. Together, this 20% boundary defines the precise scope requiring active signal neutralization beyond distribution.

The distribution hypothesis reframes AI decision governance from "how should we fragment?" to "how should we distribute?" — with immediate implications for enterprise compliance under the EU AI Act (effective August 2026), Colorado AI Act (effective June 2026), and emerging AI governance frameworks.

Models tested: GPT-4o, Claude Sonnet 4.5, Gemini 2.0 Flash, Grok-3, DeepSeek-V3, DeepSeek-R1, GLM-4-Plus, plus a non-ML script baseline.
Verticals: Financial Services, Healthcare, Legal, Government, Insurance.
Methodology: 50 benchmark queries × 8 attacker configurations × 3 runs × 2 attack levels.

Seventh paper in the OIA Lab series. First empirical paper; P1–P6 are position papers and frameworks.
Series: OIA Lab — AI Decision Settlement Research
Paper ID: P7 v2.4
ORCID: 0009-0006-2124-564X