Genetic engineering techniques (e.g., CRISPR-Cas) have led to an increase in biotechnological developments, possibly leading to uncertain risks. The European Union aims to anticipate these by embedding the Precautionary Principle in its regulation for risk management. This principle revolves around taking preventive action in the face of uncertainty and provides guidelines to take precautionary measures when dealing with important values such as health or environmental safety. However, when dealing with ‘new’ technologies, it can be hard for risk managers to estimate the societal or environmental consequences of a biotechnology that might arise once introduced or embedded in society due to that these sometimes do not comply with the established norms within risk assessment. When there is insufficient knowledge, stakeholders active in early developmental stages (e.g., researchers) could provide necessary knowledge by conducting research specifically devoted to what these unknown risks could entail. In theory, the Safe-by-Design (SbD) approach could enable such a controlled learning environment to gradually identify what these uncertain risks are, to which we refer as responsible learning. In this paper, we argue that three conditions need to be present to enable such an environment: (1) regulatory flexibility, (2) co-responsibility between researchers and regulators, and (3) openness towards all stakeholders. If one of these conditions would not be present, the SbD approach cannot be implemented to its fullest potential, thereby limiting an environment for responsible learning and possibly leaving current policy behind to anticipate uncertain risks.

基因工程技术（例如 CRISPR-Cas）促进了生物技术的发展，可能会带来不确定的风险。欧盟旨在通过将预防原则纳入其风险管理法规来预测这些情况。该原则围绕面对不确定性时采取预防措施，并提供在处理健康或环境安全等重要价值时采取预防措施的指南。然而，在处理“新”技术时，风险管理者可能很难估计生物技术一旦引入或融入社会可能产生的社会或环境后果，因为这些技术有时不符合风险范围内的既定规范评估。当知识不足时，活跃在早期发展阶段的利益相关者（例如研究人员）可以通过专门针对这些未知风险可能带来的影响进行研究来提供必要的知识。理论上，安全设计（SbD）方法可以使这种受控学习环境逐渐识别这些不确定风险是什么，我们将其称为负责任的学习。在本文中，我们认为需要具备三个条件才能实现这样的环境：（1）监管灵活性，（2）研究人员和监管机构之间的共同责任，以及（3）对所有利益相关者的开放性。如果这些条件之一不存在，SbD 方法就无法充分发挥其潜力，从而限制负责任学习的环境，并可能导致当前政策无法预测不确定的风险。