---
format: "aidr-story-markdown/v1"
id: "f5862f5288d98e8f71dfabb075b4021643f0fde2fd582d200a8f9cde7455e643"
canonical_url: "https://aidr.today/f5862f52?lang=en"
title: "Adversarial Risk in the Generative AI Era Necessitates Dropping the Small Epsilon Ball"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-11-13T05:00:00.000Z"
category: "Research"
topics: ["safety"]
source_urls: ["https://arxiv.org/abs/2611.14594"]
summary: "Adversarial Machine Learning research focuses on finding the epsilon-ball, which involves minimal perturbations designed to fool models. The research argues that fixation on $\\ell_p$ norms misleads defenders into optimizing for narrow, mathematical robustness. This paradigm may actively facilitate harm by communicating a false sense of security against the wrong threats."
---

# Adversarial Risk in the Generative AI Era Necessitates Dropping the Small Epsilon Ball

> [Open the canonical story](<https://aidr.today/f5862f52?lang=en>)

**Published:** 2026-11-13T05:00:00.000Z
**Category:** Research
**Topics:** safety

## Summary

Adversarial Machine Learning research focuses on finding the epsilon\-ball, which involves minimal perturbations designed to fool models\. The research argues that fixation on $\\ell\_p$ norms misleads defenders into optimizing for narrow, mathematical robustness\. This paradigm may actively facilitate harm by communicating a false sense of security against the wrong threats\.

## Sources

- [Story source](<https://arxiv.org/abs/2611.14594>)