Abstract. In a recent theoretical article [S.H. Kazemi, S. Ghanbari, M. Mahmoudi, Eur. Phys. J. D 70, 1 (2016)], Kazemi et al. claim to have demonstrated superluminal light transmission in an optomechanical system where a Bose-Einstein condensate serves as the mechanical oscillator. In fact the superluminal propagation is only inferred from the existence of a minimum of transmission of the system at the probe frequency. This condition is not sufficient and we show that, in all the cases where superluminal propagation is claimed by Kazemi et al., the propagation is in reality subluminal. Moreover, we point out that the system under consideration is not minimum-phase-shift. The Kramers-Kronig relations then only fix a lower limit to the group delay and we show that these two quantities have sometimes opposite signs.When the transmission of light in a given medium displays a well-marked narrow dip at some frequency, the group velocity at this frequency may be larger than the velocity of light in vacuum or even negative. An ideally smooth light-pulse can then exits the medium without significant distortion before than if it had propagated in vacuum [1]. Such superluminal or fast propagation is not at odds with relativistic causality since a given point of the output-pulse profile is not a direct reflection of the homologous point of the incident-pulse profile but results from the action of the medium on all the earlier part of the incident pulse. The major challenge in such experiments is to obtain advancements comparable to the pulse duration with moderate distortion. Convincing experiments have been performed in the 1980s [2,3] in media with a narrow absorption line. Unfortunately, superluminal propagation is then accompanied by strong absorption. This inconvenience is overcome by using a medium with a doublet of gain lines [4,5] and a minimum of transmission between them. Significant advancements have been evidenced in an atomic vapor with this arrangement [6]. A comprehensive review on fast light in atomic media can be found in [7]. Experiments involving four wave mixing are reported in [8].Superluminal or subluminal propagation can only be demonstrated by a determination of the group delay and one should not hastily conclude from what it precedes that every gain system with a dip in its transmission curve will be superluminal. This extrapolation is unfortunately made in a recent theoretical article [9] whose authors claim to have evidenced superluminal propagation by giving this a e-mail :bernard.segard@univ-lille1.fr sole argument. The system under consideration is an optomechanical device consisting in a high-quality optical cavity containing a Bose-Einstein condensate (BEC) of Rubidium atoms which serves as the mechanical oscillator [10]. It is submitted to a strong pump field (continuous wave) and to a weak probe field. In a frame rotating at the pump angular-frequency, the transfer function for the probe field reads as [9] 1withIn these expressions, κ (γ m ) is the damping rate of the cavity (the mechanic...