Artificial Intelligence (AI) is a word that has been used in many articles or revues to cover different technologies and systems. For example, the word is applied to automated cars, security systems, Roombas and bots. A bot is a device or software which can execute commands, reply to messages or perform routine tasks with minimum, if not any, human intervention.
Mid-June 2017, Facebook’s bot created its own language while negotiating with another bot. A wave of panic issued from this news whereas it is not uncommon for AI to develop its own language. A question may however arise as to whether humans are able to understand the language. If a problem were to occur, could humans understand the language and see what went wrong? How can AI be defined? What are its implications in our society? More particularly, what implications does AI have in the legal sphere?
AI can be broadly defined as a technology or system making it possible for a computer to perform tasks involving a simulation of human intelligence including decision making or learning. Such a broad definition offers flexibility when it comes to innovation and can be applied to many fields namely the medical field or the tech industries. Regarding the legal field, the definition must be narrowed to draft regulations and legislations. AI may be narrowly defined as technologies capable of processing different and various data namely unstructured data.
According to the field in which AI applies, it does not present the same challenges and risks. An automated car is not the same as surgical robot or AI technology used to research case law. Each field has its own particularities which must be taken into account when developing and regulating AI. However, it remains unclear how legislators will decide to regulate this technology.
As of today, there is no specific legal framework governing Artificial Intelligence. In the United-States, some courts have addressed the problem of liability. Other countries have also started researching how to accommodate existing laws with AI and anticipate the risks. For example, in October 2016, the European Parliament’s Legal Affairs Committee commissioned and published a study on the European civil law rules in Robotics.
The European Union must also address questions and issues regarding the coming into force of the General Data Protection Regulation of 27 April 2016 (regulation n°2016/679) on May 25th, 2018. AI operates on data and could fall under the regulation. If this is the case, the AI technology will be subject to new obligations and procedures such as the principle of transparency or the principle of data minimization.
Everyone is concerned by these issues. Indeed, AI has also had a significant impact on professions in general. The impact AI technology has on our daily lives is an interesting point to keep mind namely when trying to figure out how to regulate it.
AI based technologies and autonomous machine are rapidly developing and occupying a significant place in our society. Legal claims relating to them will not cease to rise and it is crucial that an appropriate legislative framework be put into place.
Despite these considerations, no harmonized or uniform solution has been found. Several main questions have emerged: what liability for AI? Is the developer, the user and the manufacturer liable? Can Artificial Intelligence claim the same rights as human (such as Intellectual Property)? Can the legislations in effect today apply to AI (for example, agency law, tort law, real property law)? Practitioners, professors, judges, legislators and so forth are actively studying and researching answers to these questions and many more.
Mathias Avocats will be working on these topics and will publish several articles about the above questions. Mathias Avocats has developed an expertise namely regarding new technologies and data protection.