算法合同意思表示错误的法律救济

Legal Remedies for Mistakes in Expression of Intent in Algorithm Contracts

算法合同意思表示错误的法律规制无法一概而论,应进行类型化的讨论。确定性算法发出的“意思表示”应被视为事前编码的用户意图的实现,算法只是纯粹的传输工具,在错误认定上采取一元论观点,算法本身导致的表示错误或第三方原因引起的动机错误均作为意思表示撤销的原因纳入重大误解范畴。非确定性算法发出的“意思表示”可能远超用户事前的意图框架,但由此认定其为具有意思能力的行动者的各种学说带来的相应体系后果对现行合同法规则冲击过大且并未解决责任承担问题。更为妥善的方案是通过对现有规则的扩张解释涵盖前述问题:用户在作出使用算法的决定时,应视为发出公开要约,同意受要约人以任何算法同意的条件订立对该用户具有法律约束力的合同;原则上用户承受算法行为带来的一切后果,不能主张错误救济,但如果由此产生的合同责任可能会在经济上摧毁算法用户,则例外性地允许其主张重大误解。

Concluding contracts using AI algorithms poses a challenge to the application of traditional contract law,particularly regarding remedies for mistakes in declaration of intention.The intent expressed by the deterministic algorithm should be viewed as its users prior thoughts,while the algorithm serves as a conduit tool.Users can claim serious misunderstanding remedy when unusual results appear,no matter such result caused by algorithm or the third party.The intents expressed by non-deterministic algorithm might surpass the user's prior thoughts.However,the theory of actors deemed capable of acting is a jarring surprise to current contract law and does not consider liability assumptions.To achieve a moderated liability allocation while minimizing the impact on contract law,non-deterministic algorithm can still be considered as pure tools.When its users decided to use algorithm,he or she shall be deemed to have made an open offer to consent to the offeree entering into a legally binding contract for that user on whatever terms the algorithm agrees.In principle the user suffers all the consequences of the algorithm's behavior and cannot claim the remedy for mistakes.However,if the resulting contractual liability is sufficient to destroy the algorithm user,the user is exceptionally allowed to claim serious misunderstanding remedy.