AI RESEARCH

TrEEStealer: Stealing Decision Trees via Enclave Side Channels

arXiv CS.LG

ArXi:2604.18716v1 Announce Type: cross Today, machine learning is widely applied in sensitive, security-related, and financially lucrative applications. Model extraction attacks undermine current business models where a model owner sells model access, e.g., via MLaaS APIs. Additionally, stolen models can enable powerful white-box attacks, facilitating privacy attacks on sensitive