Work

Malice in Agentland: Down the Rabbit Hole of Backdoors in the AI Supply Chain

Year: 2026

Type: conference-paper

Abstract: While finetuning AI agents on interaction data—such as web browsing or tool use—improves their capabilities, it also introduces critical security vulnerabilities within the agentic AI supply chain.

DOI: 10.1145/3786335.3813166

Authors: Léo Boisvert, Abhay Puri, Chandra Kiran Reddy Evuru, Nazanin Mohammadi Sepahvand, Nicolas Chapados, Quentin Cappart, Alexandre Lacoste, Krishnamurthy Dvijotham, Alexandre Drouin, Jason Stanley

Institutions: Mila - Quebec Artificial Intelligence Institute, Polytechnique Montréal, ServiceNow (United States), UCLouvain

Language: English

Topic: Adversarial Robustness in Machine Learning

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: gold

OpenAlex ID: W4414973802

FWCI:

Cites:

Cited by:

Related to: