Policy option 1
Policy option 1(a)
Prohibit employer monitoring of workers in particular contexts.
This could include monitoring:
a.Outside of work (temporally or geographically)
b. Physical and relational contexts ‘in work’ where data collection or monitoring poses risks to human dignity or the exercise of fundamental rights, including:
(i) in private spaces such as bathrooms; and
(ii) in private conversations or communications, especially including conversations and communications with worker representatives.
What does this mean in practice?
- Clear limits on the monitoring of workers outside of working hours and working premises, including time on breaks or off duty.
- Prohibits monitoring of employees in private spaces in the workplace that are not intrinsically connected to work.
- Forbids the monitoring of conversations with workers’ representatives in all circumstances.
Real-world examples of policy option 1(a)
Data protection law: certain EU Member States have prohibited the processing of workers’ personal data in private spaces. Expert guidance from the Article 28 Working Party also states that employees must have ‘certain private spaces to which the employer may not gain access under any circumstances’.
Labour law in New South Wales: prohibits the surveillance of employees in specific locations (e.g., changing rooms and bathrooms) and outside of work on personal devices (see Workplace Surveillance Act 2005 No 47 (New South Wales), s 15-16)
Human rights law in Europe: The European Court of Human Rights (ECtHR) in Barbulescu v Romania established a positive obligation on states to safeguard employees’ right to private life from monitoring by employers (Atkinson, 2018).
Policy option 1(b)
A prohibition of worker monitoring for specific high-risk purposes.
Prohibit monitoring and collection and processing of any data (personal or non-personal) for purposes that pose risks to human dignity and fundamental rights, including:
a. Emotional or psychological manipulation,
b. Prediction of, or persuasion against, the exercise of legal rights, especially including the right to organise; and/or
c. Other context-specific harmful purposes.
What does this mean in practice?
Forbids technology that emotionally or psychologically manipulates workers. Prohibits the processing of data (even if anonymised) that is used to prevent workers from exercising their legal rights, e.g., the right to organise.
Real-world examples of policy option 1(b)
EU AI Act: prohibits the use of AI systems to infer individuals’ emotions in the workplace and biometric categorisation systems that deduce or infer trade union membership (EU AI Act, art 5(f)-(g)).
Data protection law: restricts the processing of special-category data (e.g., personal data revealing trade union membership or biometric data) (EU GDPR, art 9(1)(2))
Labour law: forbids digital labour platforms from processing data to predict the exercise of platform workers’ fundamental rights or to infer their trade union membership (Platform Work Directive, art 7(d)-(e))
Employment law in California: prohibits employers from making employment-related decisions using automated decision systems that draw upon emotion recognition technologies (California Bill AB-1651, 2021-2022, s1543(b)(6))
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.