Model Weight Theft With Just Noise Inputs: The Curious Case of the Petulant Attacker
This reveals a vulnerability in convolutional neural networks where weight theft is possible with minimal access, highlighting security risks for deployed AI models.
The paper investigates how an attacker can steal model weights using only noise inputs and access to the softmax layer, achieving 96% test accuracy on MNIST and 82% on KMNIST with i.i.d. Bernoulli noise.
This paper explores the scenarios under which an attacker can claim that 'Noise and access to the softmax layer of the model is all you need' to steal the weights of a convolutional neural network whose architecture is already known. We were able to achieve 96% test accuracy using the stolen MNIST model and 82% accuracy using the stolen KMNIST model learned using only i.i.d. Bernoulli noise inputs. We posit that this theft-susceptibility of the weights is indicative of the complexity of the dataset and propose a new metric that captures the same. The goal of this dissemination is to not just showcase how far knowing the architecture can take you in terms of model stealing, but to also draw attention to this rather idiosyncratic weight learnability aspects of CNNs spurred by i.i.d. noise input. We also disseminate some initial results obtained with using the Ising probability distribution in lieu of the i.i.d. Bernoulli distribution.