On the robustness and security of Artificial Intelligence systems

Dmitry Namiot, Eugene Ilyushin


In the modern interpretation, artificial intelligence systems are machine learning systems. Often this is even further narrowed down to artificial neural networks. The robustness of machine learning systems has traditionally been considered as the main issue that determines the applicability of machine learning systems in critical areas (avionics, driverless movement, etc.). But is robustness alone sufficient for such applications? It is precisely this issue that this article is devoted to. Will robust systems always be reliable and safe for use in critical areas? For example, the classical definition of robustness speaks of maintaining the efficiency of the system (consistency of its conclusions) under small perturbations of the input data. But this same definition does not say anything about the correctness of the results obtained. In the classical formulation, we are talking about small (imperceptible, speaking of images) data changes, but this “smallness”, in fact, has two very specific reasons. Firstly, this corresponds precisely to the human understanding of sustainability, when small (imperceptible) changes should not affect the result. Secondly, small changes allow us to formally describe data manipulations. But if we are talking about M2M systems, then the size (degree) of data change does not matter. Robustness alone is not enough to conclude that a machine learning system is secure.

Full Text:

PDF (Russian)


Artificial Intelligence in Cybersecurity. http://master.cmc.msu.ru/?q=ru/node/3496 (in Russian) Retrieved: May, 2022.

Namiot, Dmitry, Eugene Ilyushin, and Ivan Chizhov. "Ongoing academic and industrial projects dedicated to robust machine learning." International Journal of Open Information Technologies 9.10 (2021): 35-46.

Namiot, Dmitry, Eugene Ilyushin, and Ivan Chizhov. "The rationale for working on robust machine learning." International Journal of Open Information Technologies 9.11 (2021): 68-74.

Why Robustness is not Enough for Safety and Security in Machine Learning https://towardsdatascience.com/why-robustness-is-not-enough-for-safety-and-security-in-machine-learning-1a35f6706601

[5] Borg, Markus, et al. "Safely entering the deep: A review of verification and validation for machine learning and a challenge elicitation in the automotive industry." arXiv preprint arXiv:1812.05389 (2018).

Namiot, Dmitry, Eugene Ilyushin, and Oleg Pilipenko. "On Trusted AI Platforms." International Journal of Open Information Technologies 10.7 (2022): 119-127.

Ilyushin, Eugene, Dmitry Namiot, and Ivan Chizhov. "Attacks on machine learning systems-common problems and methods." International Journal of Open Information Technologies 10.3 (2022): 17-22.

Namiot, Dmitry, Eugene Ilyushin, and Ivan Chizhov. "On a formal verification of machine learning systems." International Journal of Open Information Technologies 10.5 (2022): 30-34.

Wang, Bolun, et al. "Neural cleanse: Identifying and mitigating backdoor attacks in neural networks." 2019 IEEE Symposium on Security and Privacy (SP). IEEE, 2019.

Zhang, Yuheng, et al. "The secret revealer: Generative model-inversion attacks against deep neural networks." Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2020.

Rigaki, Maria, and Sebastian Garcia. "A survey of privacy attacks in machine learning." arXiv preprint arXiv:2007.07646 (2020).

Cool Or Creepy? Facebook Is Building An AI That Sees The World Like Humans Do https://wechoiceblogger.com/cool-or-creepy-facebook-is-building-an-ai-that-sees-the-world-like-humans-do/

Tian, Yuchi, et al. "Deeptest: Automated testing of deep-neural-network-driven autonomous cars." Proceedings of the 40th international conference on software engineering. 2018.

Allen-Zhu, Zeyuan, and Yuanzhi Li. "Feature purification: How adversarial training performs robust deep learning." 2021 IEEE 62nd Annual Symposium on Foundations of Computer Science (FOCS). IEEE, 2022.

Goodfellow, Ian, Patrick McDaniel, and Nicolas Papernot. "Making machine learning robust against adversarial inputs." Communications of the ACM 61.7 (2018): 56-66.

Dong, Guozhu, and Huan Liu, eds. Feature engineering for machine learning and data analytics. CRC Press, 2018.

Dmitry, Namiot, Ilyushin Eugene, and Chizhov Ivan. "On a formal verification of machine learning systems." International Journal of Open Information Technologies 10.5 (2022): 30-34.

Ruan, Wenjie, et al. "Global robustness evaluation of deep neural networks with provable guarantees for the hamming distance." International Joint Conferences on Artificial Intelligence Organization, 2019.

Gopinath, Divya, et al. "Deepsafe: A data-driven approach for assessing robustness of neural networks." International symposium on automated technology for verification and analysis. Springer, Cham, 2018.

Plex https://ai.googleblog.com/2022/07/towards-reliability-in-deep-learning.html

Shafaei, Sina, et al. "Uncertainty in machine learning: A safety perspective on autonomous driving." International Conference on Computer Safety, Reliability, and Security. Springer, Cham, 2018.

Francisco Herrera Dataset Shift in Classification: Approaches and Problems http://iwann.ugr.es/2011/pdf/InvitedTalk-FHerreraIWANN11.pdf Retrieved: Jul, 2022

Lu, Jie, et al. "Learning under concept drift: A review." IEEE Transactions on Knowledge and Data Engineering 31.12 (2018): 2346-2363

Fijalkow, Nathanaël, and Mohit Kumar Gupta. "Verification of neural networks: Specifying global robustness using generative models." arXiv preprint arXiv:1910.05018 (2019).

Everything You “Know” About Software and Safety is Probably Wrong https://2020.icse-conferences.org/details/icse-2020-plenary/8/Everything-You-Know-About-Software-and-Safety-is-Probably-Wrong

Identifying and eliminating bugs in learned predictive models https://www.deepmind.com/blog/identifying-and-eliminating-bugs-in-learned-predictive-models.


  • There are currently no refbacks.

Abava  Absolutech Convergent 2020

ISSN: 2307-8162