Work and automation

Additional Information

2.1 Organisations that implement AI systems in the workplace should provide opportunities for affected employees to participate in the decision-making process related to such implementation. 2.2 Consideration should be given as to whether it is achievable from a technological perspective to ensure that all possible occurrences should be pre-decided within an AI system to ensure consistent behaviour. If this is not practicable, organisations developing, deploying or using AI systems should consider at the very least the extent to which they are able to confine the decision outcomes of an AI system to a reasonable, non-aberrant range of responses, taking into account the wider context, the impact of the decision and the moral appropriateness of “weighing the unweighable” such as life vs. life. 2.3 Organisations that develop, deploy or use AI systems that have an impact on employment should conduct a Responsible AI Impact Assessment to determine the net effects of such implementation. 2.4 Governments should closely monitor the progress of AI-driven automation in order to identify the sectors of their economy where human workers are the most affected. Governments should actively solicit and monitor industry, employee and other stakeholder data and commentary regarding the impact of AI systems on the workplace and should develop an open forum for sharing experience and best practices. 2.5 Governments should promote educational policies that equip all children with the skills, knowledge and qualities required by the new economy and that promote life-long learning. 2.6 Governments should encourage the creation of opportunities for adults to learn new useful skills, especially for those displaced by automation. 2.7 Governments should study the viability and advisability of new social welfare and benefit systems to help reduce, where warranted, socio-economic inequality caused by the introduction of AI systems and robotic automation

Part of