English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Explainable Artificial Intelligence (XAI) encompasses a range of techniquesand procedures aimed at elucidating the decision-making processes of AI models.While XAI is valuable in understanding the reasoning behind AI models, the dataused for such revelations poses potential security and privacy vulnerabilities.Existing literature has identified privacy risks targeting machine learningmodels, including membership inference, model inversion, and model extractionattacks. Depending on the settings and parties involved, such attacks maytarget either the model itself or the training data used to create the model.  We have identified that tools providing XAI can particularly increase thevulnerability of model extraction attacks, which can be a significant issuewhen the owner of an AI model prefers to provide only black-box access ratherthan sharing the model parameters and architecture with other parties. Toexplore this privacy risk, we propose AUTOLYCUS, a model extraction attack thatleverages the explanations provided by popular explainable AI tools. Weparticularly focus on white-box machine learning (ML) models such as decisiontrees and logistic regression models.  We have evaluated the performance of AUTOLYCUS on 5 machine learningdatasets, in terms of the surrogate model's accuracy and its similarity to thetarget model. We observe that the proposed attack is highly effective; itrequires up to 60x fewer queries to the target model compared to thestate-of-the-art attack, while providing comparable accuracy and similarity. Wefirst validate the performance of the proposed algorithm on decision trees, andthen show its performance on logistic regression models as an indicator thatthe proposed algorithm performs well on white-box ML models in general.Finally, we show that the existing countermeasures remain ineffective for theproposed attack.

AUTOLYCUS: Exploiting Explainable AI (XAI) for Model Extraction Attacks against White-Box Models