# Estimating the Brittleness of AI: Safety Integrity Levels and the Need for Testing Out-Of-Distribution Performance

arXiv: https://arxiv.org/abs/2009.00802 · HTML: https://mlsanity.com/paper/2009.00802/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Sep 2020.

## Built on by (most cited first, 8 of 8)

- [Improved OOD Generalization via Adversarial Training and Pre-training](https://mlsanity.com/paper/2105.11144/) (2021)
- [Retrospectives on the Embodied AI Workshop](https://mlsanity.com/paper/2210.06849/) (2022)
- [Evaluating the Robustness of Collaborative Agents](https://mlsanity.com/paper/2101.05507/) (2021)
- [Breaking Correlation Shift via Conditional Invariant Regularizer](https://mlsanity.com/paper/2207.06687/) (2022)
- [Military AI Needs Technically-Informed Regulation to Safeguard AI Research and its Applications](https://mlsanity.com/paper/2505.18371/) (2025)
- [Out-of-Distribution Detection for Safety Assurance of AI and Autonomous Systems](https://mlsanity.com/paper/2510.21254/) (2025)
- [Gradient-Optimized Fuzzy Classifier: A Benchmark Study Against State-of-the-Art Models](https://mlsanity.com/paper/2504.16263/) (2025)
- [Rock, Paper, Scissors, ... Dynamite - A Model of Disruption from New Technologies](https://mlsanity.com/paper/2609.00207/) (2026)

## Abstract (arXiv)

Test, Evaluation, Verification, and Validation (TEVV) for Artificial Intelligence (AI) is a challenge that threatens to limit the economic and societal rewards that AI researchers have devoted themselves to producing. A central task of TEVV for AI is estimating brittleness, where brittleness implies that the system functions well within some bounds and poorly outside of those bounds. This paper argues that neither of those criteria are certain of Deep Neural Networks. First, highly touted AI successes (eg. image classification and speech recognition) are orders of magnitude more failure-prone than are typically certified in critical systems even within design bounds (perfectly in-distribution sampling). Second, performance falls off only gradually as inputs become further Out-Of-Distribution (OOD). Enhanced emphasis is needed on designing systems that are resilient despite failure-prone AI components as well as on evaluating and improving OOD performance in order to get AI to where it can clear the challenging hurdles of TEVV and certification.