Think about smart fridges, shopping bots, auction bots or virtual digital assistants such as Amazon Echo: consumers increasingly make use of smart devices both to buy and sell everyday products. All of these commercial applications assist users in negotiating and concluding contracts.
At first sight, the fact that the user configures and activates the device is not in doubt compared to his/her willingness to conclude transactions. From this perspective, smart devices are not substantially different from computers or faxes. However, intelligent agents differ from other electronic means of communication by their self-learning capacity and their high degree of autonomy. As devices can act without human intervention, this means that the user does not know in advance what the legal outcome of such action by the device will be. Consequently, it is likely that the device enters into a contract that its user or owner could not reasonably expect or did not want.
It is therefore important to determine when and under what conditions contracts concluded by autonomous devices are legally binding for their user or owner. In the absence of such legal certainty, there is a risk that the device user or owner systematically tries to invoke the nullity of the transaction by arguing that his/her consent was lacking. This issue is particularly important as the European e-Commerce Directive stipulates that Member States must ensure that their legal systems allow contracts to be concluded electronically.
As far as consumers are concerned (B2C), they can exercise their right of withdrawal as a last resort, even if the contract is validly concluded via an autonomous device. Professionals (B2B), by contrast, do not have such a right of withdrawal and cannot therefore escape the undesirable consequences of such an autonomous contract conclusion.
The current regulatory regime (more precisely the European and Belgian rules governing distance and electronic contracts) is only partly capable of dealing with the contracts concluded by smart agents. As these agents develop and become more autonomous, there is no doubt that this issue will become increasingly important in the coming years. National legislators must therefore be aware of the growing forms of autonomous contracting and seek solutions to ‘bridge the gap’ between our traditional legal concepts and current and future technology.
In this respect, the solution adopted in the United States could be a valuable source of inspiration. Not only does U.S. Federal law (UETA and UCITA) recognise the validity of agreements concluded autonomously, it also presumes the binding character of contracts performed by a smart agent for its user or owner.