In anticipation of the advent of autonomous shipping, it is contended that the present fault-based collision liability regimes of most jurisdictions do not sufficiently cater for the situations that may arise.
The wording of the 1910 Collison Convention and the national collision regimes based on it (often applicable also outside the rather narrow scope of application of the convention itself) assume that fault can either be found among the vessels (read: the shipowners or their servants) themselves or not at all. The latter cases have traditionally been referred to as events outside of the ship.
The wording of the convention does not cater for the situation that the collision was caused by the fault of a third-party contractor in relation to work undertaken to one of the ships, despite the fact that in such a situation, the cause must be said to rather lie “within” that ship than outside of it. At the same time, in relation to autonomous shipping, third party contractors, such as software or internet providers, are expected to play an increasingly important role for the safety and security of the ship.
It is unclear if a national court faced with a collision caused by the fault of such a contractor would apply Art. 4 of the convention by analogy. The provision deals with the situation that both vessels are at fault but that the respective degree cannot be established. The result is a 50/50 liability split. Alternatively, the court may instead draw an analogy to Art. 2, dealing with the situation that no vessel was at fault. In such a case each vessel has to carry its own damage. A third viable option is that the court would consider the situation to fall outside of the scope of the convention all together and apply instead to other national rules. The difference may be substantial.
Also, the traditional means of establishing fault will be affected by how the human actions or omissions will move further away from the collision in time and space. Not only can this be expected to give rise to complex questions of causation, it can also be expected to affect the statutory rules serving as guidance for correct action. For example, without human involvement in the navigation, a breach of the rather straightforward “dos” or “don’ts” of COLREG (Convention on the International Regulations for Preventing Collisions at Sea) can no longer serve as indication for negligence in navigation. Rather, one must look at negligence in management of the ship with guidance in the more generally worded international and national objectives and principles. This arguably increases the courts’ leeway to affirm or reject fault in the individual case.
In essence, it is contended that the current fault-based collision liability regime applied to autonomous ships does not preserve the unification and foreseeability that the 1910 Collison Convention was once drafted to guarantee.