In this paper a novel method is introduced for target detection in bistatic passive radars which uses the concept of correntropy to distinguish correct targets from false detections. In proposed method the history of each cell of ambiguity function is modeled as a stochastic process. Then the stochastic processes consist the noise are differentiated from those consisting targets by constructing an FIR adaptive filter. A cost function which is based on correntropy is utilized to update the filter. The performance of the proposed method is evaluated by simulation in presence of rapid and slow moving targets. The obtained results shows the superiority of the proposed method compared to its alternatives in such manner that it detects rapid targets at least 18.7 and 20.1 percent better than HOSCM and PFCM. Furthermore it detects slow targets 19.3 and 21.4 percent better than those alternatives, all in presence of maximum noise (i.e. SCNR=-30dB).