GCD AI RADAR
incidents/AIR-2023-0012

Mobile Deep Learning Apps Vulnerable to Neural Payload Injection

Attack on AI1 Research / proof of conceptData poisoningSupply chain

Researchers found that many deep learning models in Google Play apps are vulnerable to backdoor attacks via neural payload injection. The empirical study identified multiple critical applications affected by this vulnerability.

Details

type
Attack on AI
An AI model, agent or AI-powered product was the thing exploited, abused or leaked.
categories
  • Data poisoningprimary
    Training or retrieval data manipulated to change model behaviour.
  • Supply chain
    Malicious or backdoored models, poisoned packages, unsafe deserialization and other compromised AI dependencies.
owasp llm top 10
mitre atlas
Both rows are derived from the type and categories (OWASP 2025 edition, ATLAS release 2026.09), not classified per record. How
severity
1 · Research / proof of concept
Demonstrated by researchers; no known real-world exploitation.
targets
Open-weight model, Enterprise deployment
products
Mobile Deep Learning Apps
incident date
31 Mar 2023
first reported
31 Mar 2023
origin
AVID
classifier confidence
0.90 · published at or above 0.80; see the methodology
tags
backdoor · data_poisoning · mobile_apps
record
created 2026-10-18 06:56Z · updated 2026-10-18 06:56Z · json

Sources · 3

  1. 2023-03-31
    AVID-2023-V013: Deep learning models are increasingly used in mobile applications as critical components.
    AVID · github.com/avidml/avid-db/blob/main/vulnerabilities/2023/AVID-2023-V013.json
  2. 2023-03-31
    Backdoor Attack on Deep Learning Models in Mobile Apps
    MITRE ATLAS · atlas.mitre.org/studies/AML.CS0013
  3. 2023-03-31
    DeepPayload: Black-box Backdoor Attack on Deep Learning Models through Neural Payload Injection
    arXiv · arxiv.org/abs/2101.06896

Titles and links belong to their publishers. The summary above is written by GCD AI RADAR in its own words; data is CC BY-SA 4.0.

Related incidents