closedclosed 14 Apr 2023Research1 Research / proof of conceptAI
CNN-Based Botnet DGA Detector Bypassed via Domain Mutation
Palo Alto Networks DGA Detector
Researchers bypassed a convolutional neural network-based botnet domain generation algorithm detector using a generic domain name mutation technique. The method successfully evaded the ML-based detection module in testing.
- first report
- 31 Mar 2023
- last update
- 31 Mar 2023
Timeline · 4
- 31 Mar 2023AVID-2023-V002: The Palo Alto Networks Security AI research team was able to bypass a Convolutional Neural Network based botnet Domain Generation Algorithm (DGA) detector using a generic domain name mutation technique.
- 31 Mar 2023Botnet Domain Generation Algorithm (DGA) Detection Evasion
- 31 Mar 2023Yu, Bin, Jie Pan, Jiaming Hu, Anderson Nascimento, and Martine De Cock. "Character level based detection of DGA domain names." In 2018 International Joint Conference on Neural Networks (IJCNN), pp. 1-8. IEEE, 2018.
- 31 Mar 2023Degas source code
Sources · 4
News and write-ups
- 2023-03-31Yu, Bin, Jie Pan, Jiaming Hu, Anderson Nascimento, and Martine De Cock. "Character level based detection of DGA domain names." In 2018 International Joint Conference on Neural Networks (IJCNN), pp. 1-8. IEEE, 2018.faculty.washington.edu · faculty.washington.edu/mdecock/papers/byu2018a.pdf
- 2023-03-31Degas source codeGitHub · github.com/matthoffman/degas
Research and catalogues
- 2023-03-31AVID-2023-V002: The Palo Alto Networks Security AI research team was able to bypass a Convolutional Neural Network based botnet Domain Generation Algorithm (DGA) detector using a generic domain name mutation technique.AVID · github.com/avidml/avid-db/blob/main/vulnerabilities/2023/AVID-2023-V002.json
- 2023-03-31Botnet Domain Generation Algorithm (DGA) Detection EvasionMITRE ATLAS · atlas.mitre.org/studies/AML.CS0001
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 31 Mar 2023 · first reported 31 Mar 2023
- origin
- AVID
- record
- confidence 0.90 · created 2026-10-18 06:56Z · updated 2026-10-18 06:56Z
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
# CNN-Based Botnet DGA Detector Bypassed via Domain Mutation https://gcdradar.vercel.app/incidents/cnn-based-botnet-dga-detector-bypassed-via-domain-mutation · AIR-2023-0013 · closed 2023-04-14 - Lane: Research - Severity: 1 (Research / proof of concept) - Fix status: Fix status unknown - First reported: 2023-03-31 - AI role: Attack on AI - Products: Palo Alto Networks DGA Detector ## Summary Researchers bypassed a convolutional neural network-based botnet domain generation algorithm detector using a generic domain name mutation technique. The method successfully evaded the ML-based detection module in testing. ## What to do No action recorded; read the sources below and the vendor's advisory. ## Latest updates - 2023-03-31: Degas source code (https://github.com/matthoffman/degas) - 2023-03-31: Yu, Bin, Jie Pan, Jiaming Hu, Anderson Nascimento, and Martine De Cock. "Character level based detection of DGA domain names." In 2018 International Joint Conference on Neural Networks (IJCNN), pp. 1-8. IEEE, 2018. (https://faculty.washington.edu/mdecock/papers/byu2018a.pdf) - 2023-03-31: Botnet Domain Generation Algorithm (DGA) Detection Evasion (https://atlas.mitre.org/studies/AML.CS0001) - 2023-03-31: AVID-2023-V002: The Palo Alto Networks Security AI research team was able to bypass a Convolutional Neural Network based botnet Domain Generation Algorithm (DGA) detector using a generic domain name mutation technique. (https://github.com/avidml/avid-db/blob/main/vulnerabilities/2023/AVID-2023-V002.json) ## Sources - 2023-03-31 AVID: AVID-2023-V002: The Palo Alto Networks Security AI research team was able to bypass a Convolutional Neural Network based botnet Domain Generation Algorithm (DGA) detector using a generic domain name mutation technique. https://github.com/avidml/avid-db/blob/main/vulnerabilities/2023/AVID-2023-V002.json - 2023-03-31 MITRE ATLAS: Botnet Domain Generation Algorithm (DGA) Detection Evasion https://atlas.mitre.org/studies/AML.CS0001 - 2023-03-31 faculty.washington.edu: Yu, Bin, Jie Pan, Jiaming Hu, Anderson Nascimento, and Martine De Cock. "Character level based detection of DGA domain names." In 2018 International Joint Conference on Neural Networks (IJCNN), pp. 1-8. IEEE, 2018. https://faculty.washington.edu/mdecock/papers/byu2018a.pdf - 2023-03-31 GitHub: Degas source code https://github.com/matthoffman/degas Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.
Related
- closedCommercial Face Identification Service MITRE AI Red Team demonstrates physical evasion attack on face identification
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 - closedShanghai government tax office facial recognition service Camera hijack attack evades facial recognition in tax system fraud
Breaches & attacks · Real loss · since 2018 (approx.)
AI - closedHive Box Facial-Recognition Locks Hive Box facial-recognition delivery lockers bypassed with printed photos
Research · since 23 Dec 2022
AI