Adversarial Evasion of Kaspersky Antimalware Neural Networks
Kaspersky Antimalware ML
Researchers exploited feature knowledge under a gray-box scenario to successfully evade detection for adversarially modified malware files. The work demonstrated how feature knowledge alone is sufficient for adversarial attacks on security ML models.
- first report
- 3 Feb 2022
- last update
- 3 Feb 2022
Timeline · 2
Sources · 2
News and write-ups
- 2022-02-03Article, "How to confuse antimalware neural networks. Adversarial attacks and protection"Kaspersky Securelist · securelist.com/how-to-confuse-antimalware-neural-networks-adversarial-attacks-an
Research and catalogues
- 2022-02-03Confusing Antimalware Neural NetworksMITRE ATLAS · atlas.mitre.org/studies/AML.CS0014
Titles and links belong to their publishers. The summary above is written by gcdRadar in its own words; data is CC BY-SA 4.0.
Classification
- lane
- Research
- AI role
- Attack on AI
- categories
- Adversarial evasion
- severity
- 1 · Research / proof of concept
- targets
- Enterprise deployment
- owasp llm top 10
- no entry for these categories
- dates
- happened 23 Jun 2021 · first reported 3 Feb 2022
- origin
- MITRE ATLASDerived from MITRE ATLAS (Apache 2.0, © MITRE).
- record
- confidence 0.95 · created 2026-10-18 06:55Z · updated 2026-10-18 06:55Z
Lane, AI role, categories and severity come from the classifier; the OWASP and ATLAS rows are derived from the categories. How these are assigned
Copy a brief
This record as a short Markdown brief, for a ticket, a chat or a coding agent.
the brief as text
# Adversarial Evasion of Kaspersky Antimalware Neural Networks https://gcdradar.vercel.app/incidents/adversarial-evasion-of-kaspersky-antimalware-neural-networks · AIR-2022-0030 · closed 2022-02-17 - Lane: Research - Severity: 1 (Research / proof of concept) - Fix status: Fix status unknown - First reported: 2022-02-03 (happened 2021-06-23) - AI role: Attack on AI - Products: Kaspersky Antimalware ML ## Summary Researchers exploited feature knowledge under a gray-box scenario to successfully evade detection for adversarially modified malware files. The work demonstrated how feature knowledge alone is sufficient for adversarial attacks on security ML models. ## What to do No action recorded; read the sources below and the vendor's advisory. ## Latest updates - 2022-02-03: Article, "How to confuse antimalware neural networks. Adversarial attacks and protection" (https://securelist.com/how-to-confuse-antimalware-neural-networks-adversarial-attacks-and-protection/102949) - 2022-02-03: Confusing Antimalware Neural Networks (https://atlas.mitre.org/studies/AML.CS0014) ## Sources - 2022-02-03 MITRE ATLAS: Confusing Antimalware Neural Networks https://atlas.mitre.org/studies/AML.CS0014 - 2022-02-03 Kaspersky Securelist: Article, "How to confuse antimalware neural networks. Adversarial attacks and protection" https://securelist.com/how-to-confuse-antimalware-neural-networks-adversarial-attacks-and-protection/102949 Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.
Related
- closedMobile deep learning apps Backdoor Attacks on Deep Learning Models in Mobile Apps
Research · since 18 Jan 2021
AI - closedCommercial Face Identification Service MITRE AI Red Team demonstrates physical evasion attack on face identification
Research · since 31 Mar 2023
AI - closedPalo Alto Networks DGA Detector CNN-Based Botnet DGA Detector Bypassed via Domain Mutation
Research · since 31 Mar 2023
AI - closedKaspersky Antimalware ML Model Kaspersky Antimalware ML Model Evaded via Feature Knowledge
Research · since 31 Mar 2023
AI - closedGoogle Translate Researchers replicate machine translation services and transfer adversarial attacks
Research · since 31 Mar 2023
AI - closedPalo Alto Networks C&C Traffic Detector Palo Alto Networks security team demonstrates evasion of deep learning malware detector
Research · since 31 Mar 2023
AI