A Content-Based Framework for Cybersecurity Refusal Decisions in Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Linder, Noa, Segal, Meirav, Antverg, Omer, Gekker, Gil, Fichman, Tomer, Bodenheimer, Omri, Maor, Edan, Nevo, Omer
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911455114166272
author Linder, Noa
Segal, Meirav
Antverg, Omer
Gekker, Gil
Fichman, Tomer
Bodenheimer, Omri
Maor, Edan
Nevo, Omer
author_facet Linder, Noa
Segal, Meirav
Antverg, Omer
Gekker, Gil
Fichman, Tomer
Bodenheimer, Omri
Maor, Edan
Nevo, Omer
contents Large language models and LLM-based agents are increasingly used for cybersecurity tasks that are inherently dual-use. Existing approaches to refusal, spanning academic policy frameworks and commercially deployed systems, often rely on broad topic-based bans or offensive-focused taxonomies. As a result, they can yield inconsistent decisions, over-restrict legitimate defenders, and behave brittlely under obfuscation or request segmentation. We argue that effective refusal requires explicitly modeling the trade-off between offensive risk and defensive benefit, rather than relying solely on intent or offensive classification. In this paper, we introduce a content-based framework for designing and auditing cyber refusal policies that makes offense-defense tradeoffs explicit. The framework characterizes requests along five dimensions: Offensive Action Contribution, Offensive Risk, Technical Complexity, Defensive Benefit, and Expected Frequency for Legitimate Users, grounded in the technical substance of the request rather than stated intent. We demonstrate that this content-grounded approach resolves inconsistencies in current frontier model behavior and allows organizations to construct tunable, risk-aware refusal policies.
format Preprint
id arxiv_https___arxiv_org_abs_2602_15689
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle A Content-Based Framework for Cybersecurity Refusal Decisions in Large Language Models
Linder, Noa
Segal, Meirav
Antverg, Omer
Gekker, Gil
Fichman, Tomer
Bodenheimer, Omri
Maor, Edan
Nevo, Omer
Computation and Language
Artificial Intelligence
Cryptography and Security
Large language models and LLM-based agents are increasingly used for cybersecurity tasks that are inherently dual-use. Existing approaches to refusal, spanning academic policy frameworks and commercially deployed systems, often rely on broad topic-based bans or offensive-focused taxonomies. As a result, they can yield inconsistent decisions, over-restrict legitimate defenders, and behave brittlely under obfuscation or request segmentation. We argue that effective refusal requires explicitly modeling the trade-off between offensive risk and defensive benefit, rather than relying solely on intent or offensive classification. In this paper, we introduce a content-based framework for designing and auditing cyber refusal policies that makes offense-defense tradeoffs explicit. The framework characterizes requests along five dimensions: Offensive Action Contribution, Offensive Risk, Technical Complexity, Defensive Benefit, and Expected Frequency for Legitimate Users, grounded in the technical substance of the request rather than stated intent. We demonstrate that this content-grounded approach resolves inconsistencies in current frontier model behavior and allows organizations to construct tunable, risk-aware refusal policies.
title A Content-Based Framework for Cybersecurity Refusal Decisions in Large Language Models
topic Computation and Language
Artificial Intelligence
Cryptography and Security
url https://arxiv.org/abs/2602.15689