智能合约算法错误的法律挑战

Algorithmic Errors in Smart Contracts: A Legal Challenge

摘要: 传统合同法规则中，合同作为双方当事人合意的产物，一方提出要约，另一方经过慎重考虑后需做出承诺，而人工智能技术的快速发展则使得算法能在当事人未参与的情况下形成合同。这也将以当事人达成协议为前提的传统合同法规则延伸到了一个新的领域。具体来讲，程序化交易中的客观事实可能因为复杂的计算机技术而无法百分百还原，当确定性算法对交易进行独立的决策调整时，是否可以适用认识错误理论，法律上又该如何分配因错误执行的智能合约所造成的损失。

Abstract: Under traditional contract law, a contract is the product of mutual consent between parties, whereby one party makes an offer, the other party makes an acceptance after careful consideration. The rapid development of artificial intelligence technology makes it possible for algorithms to form contracts without the participation of the parties. This extends the traditional rules of contract law based on the agreement of the parties to a new field. Specifically, the objective facts in programmed transactions may not be fully reconstructed because of the complexity of computer technology. Can the theory of cognitive error be applied when the deterministic algorithm adjusts the transaction independently? How should the law allocate the losses caused by the wrongly executed smart contracts?