While most countries have not given any thought yet to regulations governing artificial intelligence, and the European Union is only at the stage of adopting resolutions recognising the need to examine this topic, Estonia is not only drafting its own legal act but wants the law to be exceptionally innovative.
The country’s national digital adviser, Marten Kaevats, says that Estonia wants the regulations to be prepared for the potential growth in AI over the longest possible timeframe. The main legal means to meet this ambitious goal is to be awarding legal personality to AI.
Giving legal personality to artificial intelligence would be accompanied by the introduction of insurance regulations enabling the creation of policies covering the liability of such creatures. Adjustments to key institutions of private law, such as the notion of declarations of will, are also being considered. Perhaps Estonian robots will obtain not only legal capacity, but also the capacity to conclude legal transactions. Kaevats is even spinning visions of companies whose only shareholders are AI entities.
Along with entitlements for AI, there would be restrictions. The Estonian act would set boundaries which AI algorithms cannot cross. This means introducing into this innovative field the core ethical principles on which Estonia bases the construction of its modern information society. Considering the unpredictability of the decision-making pathways taken by self-teaching algorithms, the question is whether they will be capable of complying with the guidelines, and if they violate them, how they will be held responsible.