Policy option 5
Prohibit automated termination of the employment relationship, and establish a requirement for meaningful human involvement in termination decisions.
What does this mean in practice?
- Prohibits automated decisions for termination.
- Requires a ‘human in the loop’ for termination decisions.
Real-world example of policy option 5
Data protection law: prohibition on automated decision-making with legal or similarly significant effects (except in certain circumstances, e.g., performance of or entering into a contract) (EU GDPR, art 22(1)). In SCHUFA, the Court of Justice of the European Union determined that a ‘human in the loop’ drawing strongly on an automated output to make a decision still fell within the scope of an ‘automated decision’.
AI Act: providers must implement measures to ensure effective oversight of high-risk systems, e.g., AI systems intended to be used in the context of employment, worker management and access to self-employment (EU AI Act, art 14).
Labour law in California: employers cannot rely solely on worker data collected through electronic monitoring to make decisions about termination, hiring, promotion, or disciplinary decisions (California Bill AB-1651, 2021-2022, s1544).
This site is part of a project that has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (Grant agreement No. 947806).
If you have any questions or feedback about this site or our research, please get in touch by emailing ai.work@law.ox.ac.uk.