Covert Attacks on Machine Learning Training in Passively Secure MPC
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866908376053579776 |
|---|---|
| author | Jagielski, Matthew Escudero, Daniel Rachuri, Rahul Scholl, Peter |
| author_facet | Jagielski, Matthew Escudero, Daniel Rachuri, Rahul Scholl, Peter |
| contents | Secure multiparty computation (MPC) allows data owners to train machine learning models on combined data while keeping the underlying training data private. The MPC threat model either considers an adversary who passively corrupts some parties without affecting their overall behavior, or an adversary who actively modifies the behavior of corrupt parties. It has been argued that in some settings, active security is not a major concern, partly because of the potential risk of reputation loss if a party is detected cheating.
In this work we show explicit, simple, and effective attacks that an active adversary can run on existing passively secure MPC training protocols, while keeping essentially zero risk of the attack being detected. The attacks we show can compromise both the integrity and privacy of the model, including attacks reconstructing exact training data. Our results challenge the belief that a threat model that does not include malicious behavior by the involved parties may be reasonable in the context of PPML, motivating the use of actively secure protocols for training. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_17092 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Covert Attacks on Machine Learning Training in Passively Secure MPC Jagielski, Matthew Escudero, Daniel Rachuri, Rahul Scholl, Peter Cryptography and Security Machine Learning Secure multiparty computation (MPC) allows data owners to train machine learning models on combined data while keeping the underlying training data private. The MPC threat model either considers an adversary who passively corrupts some parties without affecting their overall behavior, or an adversary who actively modifies the behavior of corrupt parties. It has been argued that in some settings, active security is not a major concern, partly because of the potential risk of reputation loss if a party is detected cheating. In this work we show explicit, simple, and effective attacks that an active adversary can run on existing passively secure MPC training protocols, while keeping essentially zero risk of the attack being detected. The attacks we show can compromise both the integrity and privacy of the model, including attacks reconstructing exact training data. Our results challenge the belief that a threat model that does not include malicious behavior by the involved parties may be reasonable in the context of PPML, motivating the use of actively secure protocols for training. |
| title | Covert Attacks on Machine Learning Training in Passively Secure MPC |
| topic | Cryptography and Security Machine Learning |
| url | https://arxiv.org/abs/2505.17092 |