As algorithmic pricing becomes more widespread, the discussion about the extent to which the use of algorithms results in an increase of collusion also intensifies. While some scholars argue that algorithms are able to collude on their own (algorithmic collusion), others claim that only the use of code to enforce collusion (collusion by code) is a serious threat. In this paper, we discuss both scenarios as well as the conditions under which collusion is likely to occur. As detection and prosecution seems rather challenging, we also discuss possible remedies. These include statistical analyses of market data, an increase in trained staff for competition authorities or even a general ban of specific classes of pricing algorithms. While current competition law seems to be prepared to tackle current issues, it might be adapted for possible future challenges, in case that autonomous algorithms become greater concerns in the future.