Download Privacy Needle App

Type to search

Cybersecurity

Researchers Reveal ‘Sleeper Agent’ Tactic for Hiding Malware in AI Models

Share
A conceptual visualisation of a neural network with glowing nodes and connections representing AI weights.

New research has detailed how malicious actors can embed functional malware payloads inside the weights of open-weight artificial intelligence (AI) models. By targeting a tiny fraction of a neural network’s parameters, attackers can create “sleeper agents” that remain dormant until triggered by specific user inputs.

The technique, recently explored by researchers at Microsoft under the name “Lazarus,” demonstrates that a malicious command can be hidden in fewer than 10 dormant neurons. These neurons are specifically selected because they have little to no effect on the model’s standard functioning, allowing the poisoned model to perform normally during traditional validation and testing.

This discovery highlights a growing risk for organisations that rely on public repositories such as Hugging Face to download pre-trained models. While open-weight models are popular for their transparency and cost-effectiveness, they are rarely inspected at the weight level, where payloads can lurk undetected.

How Sleeper Agents Bypass Detection

The complexity of modern large language models (LLMs) provides significant cover for these payloads. A model with billions of parameters offers an immense search space; if an attacker modifies only 0.00001% of the network, the change is virtually impossible to spot through manual inspection or standard performance benchmarks.

The “Lazarus” method involves encoding a payload into neurons that are configured to fire sequentially only when a specific trigger is met. This trigger could be a particular phrase from a user, a specific date, or even a command read from a terminal. Until that trigger is present, the model behaves exactly as expected, effectively acting as a sleeper agent within a corporate environment.

This research builds on earlier concepts like “MaleficNet,” developed by researchers including Briland Hitaj, a senior AI security research engineer at Confidencial. That technique involved fragmenting malware binaries and spreading the pieces across millions of weights. Because the weights appear as mere noise to anyone inspecting the file, the binary remains hidden until a decoder mechanism recombines and executes it.

Defending Against AI Model Poisoning

Security experts advise that defending against weight-based poisoning requires a shift in how organisations treat AI files. Rather than viewing them as static data, they should be treated with the same scrutiny as software binaries or source code.

Luigi Vincenzo Mancini, a professor at Sapienza University of Rome, suggests a “defence-in-depth” strategy. One effective control is “diffing” the deployed weights against the original, unmodified weights from a trusted vendor or official source. This can reveal subtle changes that might indicate tampering.

Furthermore, because no single scan can guarantee a model is clean, operational controls are essential. Security teams should monitor for unexpected behaviour at the system boundary, such as an AI agent suddenly emitting shell commands, attempting unauthorised network connections, or accessing credentials it does not require for its primary task.

To mitigate the risk of a triggered sleeper agent, organisations are encouraged to restrict the access granted to AI agents, log all autonomous actions, and ensure that human approval is required for any high-consequence system changes.

Watch Our Latest Video
Stay ahead with expert insights on privacy, cybersecurity, artificial intelligence, data protection and compliance.
Privacy Needle AI | AI DPO Assistant & NDPA Compliance Software
Published: November 2, 2026
Daily Privacy News
Cybersecurity Updates
Data Protection Tips
GDPR & NDPA Explained
Tags:
Ikeh James Certified Data Protection Officer (CDPO) | NDPC-Accredited

Ikeh James Ifeanyichukwu is a Certified Data Protection Officer (CDPO) accredited by the Institute of Information Management (IIM) in collaboration with the Nigeria Data Protection Commission (NDPC). With years of experience supporting organizations in data protection compliance, privacy risk management, and NDPA implementation, he is committed to advancing responsible data governance and building digital trust in Africa and beyond. In addition to his privacy and compliance expertise, James is a Certified IT Expert, Data Analyst, and Web Developer, with proven skills in programming, digital marketing, and cybersecurity awareness. He has a background in Statistics (Yabatech) and has earned multiple certifications in Python, PHP, SEO, Digital Marketing, and Information Security from recognized local and international institutions. James has been recognized for his contributions to technology and data protection, including the Best Employee Award at DKIPPI (2021) and the Outstanding Student Award at GIZ/LSETF Skills & Mentorship Training (2019). At Privacy Needle, he leverages his diverse expertise to break down complex data privacy and cybersecurity issues into clear, actionable insights for businesses, professionals, and individuals navigating today’s digital world.

  • 1

You Might also Like

Leave a Reply

Your email address will not be published. Required fields are marked *

  • Rating

This site uses Akismet to reduce spam. Learn how your comment data is processed.