In this work, we demonstrate universal multi-party poisoning attacks that adapt and apply to any multi-party learning process with arbitrary interaction pattern between the parties. More generally, we introduce and study $(k,p)$-poisoning attacks in which an adversary controls $k\in[m]$ of the parties, and for each corrupted party $P_i$, the adversary submits some poisoned data $\mathcal{T}'_i$ on behalf of $P_i$ that is still ``$(1-p)$-close'' to the correct data $\mathcal{T}_i$ (e.g., $1-p$ fraction of $\mathcal{T}'_i$ is still honestly generated). We prove that for any ``bad'' property $B$ of the final trained hypothesis $h$ (e.g., $h$ failing on a particular test example or having ``large'' risk) that has an arbitrarily small constant probability of happening without the attack, there always is a $(k,p)$-poisoning attack that increases the probability of $B$ from $\mu$ to by $\mu^{1-p \cdot k/m} = \mu + \Omega(p \cdot k/m)$. Our attack only uses clean labels, and it is online. More generally, we prove that for any bounded function $f(x_1,\dots,x_n) \in [0,1]$ defined over an $n$-step random process $\mathbf{X} = (x_1,\dots,x_n)$, an adversary who can override each of the $n$ blocks with even dependent probability $p$ can increase the expected output by at least $\Omega(p \cdot \mathrm{Var}[f(\mathbf{x})])$.
翻译:在这项工作中,我们展示了普世多方中毒袭击,这些袭击适应和适用于任何多方学习过程,并带有各方之间任意互动的模式。更一般地说,我们介绍和研究一个对手控制了缔约方美元[m]美元[m]美元]的美元(k,p)的溢价袭击。对于每个腐败的一方,对手代表美元(P) 来提交一些中毒数据 $\mathcal{T}_i$(f) $-p$($-p),该美元($-c) 用于正确的数据 $\mac{T} 美元(g.k,p) 美元(x) 美元(x) 美元(n) 美元(x) 美元(x) 美元(x) 美元(x) 美元(x) 美元(x) 美元(x) 美元(x) 美元(m) 美元(x) 美元(x) 美元(x) 美元(x) 美元/美元(x) 美元(x) 美元(x) 美元(x) 美元(x) 美元(x) 美元(x(x(x) 美元/美元) 美元(x)的直值(x) (x(x) (l) 美元) (美元) 美元) 美元) (美元) (美元/(x(x(x) (美元) 美元) 美元) (美元) (美元) (美元) (美元/(美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元/(美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元) (美元/(美元) (美元) (美元) (美元) (美元) (美元) (美元