In both the EU and US, there are safe harbour provisions in place that should, under certain conditions, provide online service providers (OSPs) with immunity in the case of intellectual property infringement by third parties. However, the significant litigation against OSPs demonstrates that the safe harbour provisions are neither effective nor efficient. By providing OSPs with immunity against third party liability, safe harbour provisions contribute to a climate where the behaviour of OSPs is dominated by short-term business interests which are conducive neither to the enforcement of intellectual property rights by the OSPs nor to legal certainty for proprietors, internet users and OSPs alike. The precondition for invoking safe harbour provisions, that one remain passive and only act reactively, leads to wilful blindness, although OSPs are best positioned to filter infringing use of content proactively. This article therefore asserts that the safe harbour provisions must be replaced by strict intermediary liability. As will be pointed out below, this transition is not as dramatic as it seems. Safe harbours provisions were drafted at a moment when OSPs, as social media, still needed to be developed. They do not protect proprietors against infringement. Moreover, the protection of OSPs against liability is an illusion. If one extrapolates the development in filter technology one can see that advocating safe harbour provisions has become a rearguard battle and that implementation of strict liability for OSPs is inescapable.