Submission 221
Centre for Future Work
Submitter group: Research institutes.
An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.
Summary
Read the original submission for the full text.
Argues that AI-based algorithmic management is degrading job quality, health and safety, privacy and worker rights, and that existing workplace, privacy and safety laws have significant gaps. It recommends tailored, employment-specific regulation (new or amended law), with the EU Platform Work Directive as a reference, covering limits, transparency and worker involvement. It describes evidence on efficiency benefits as mixed.
Key points
Quotes are the submitter’s own words. Text taken from documents, including scanned pages read by text-recognition software, can differ from the PDF. Check the original. Listed in the same order as on the submission’s page in the interactive guide.
Argues that algorithmic management is downgrading working conditions, creating health and safety risks, perpetuating bias and undermining workers' dignity, privacy, voice and rights to organise.
AM is downgrading working conditions, creating health and safety risks, perpetuating bias and discrimination, undermining workers’ dignity and infringing on rights to privacy. It is changing workplace relationships and restricting workers’ voice, and their rights to organise.
Argues that pervasive surveillance, work intensification and loss of privacy under algorithmic management have significant consequences for work health and safety and mental health.
Pervasive and intrusive surveillance, work intensification, and loss of privacy have significant consequences for work health and safety, and impact on mental health.
Argues that there are significant shortcomings and gaps in existing regulation of algorithmic management in Australia, requiring legislative reform.
In Australia, there are significant shortcomings and gaps in existing regulations in relation to AM practices, requiring legislative reform.
Argues that reliance on education and information is not a solution to long-standing poor compliance with existing laws, and that algorithmic management presents new challenges for their effectiveness.
Reliance on education and information to address long-standing problems of poor compliance with, for example, requirements for employers to genuinely consult with workers, is not a solution. AM presents new challenges for the effectiveness of existing regulations.
Says the EU Platform Work Directive provides a key reference point for shaping regulation of algorithmic management in Australia.
The European Union’s Platform Work Directive provides a key reference point to guide the shaping of regulation of AM practices in Australia.
Says regulation of algorithmic management could be a single new law, adaptation of existing workplace laws, or a mix of the two.
The regulation of AM at work could be achieved a single piece of legislation, the adaptation of existing legislation such as workplace laws, or a mix of the two.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Coordinate policy interventions and invest significantly in skills to preserve accessible quality jobs, including entry-level jobs.
Ensuring this will require coordinated policy interventions and significant investments in skills so that quality jobs, including entry-level jobs, are preserved and accessible.
Establish clear limits on automated monitoring and decision-making, including limits and some prohibitions on collecting and using personal data.
Regulation should establish clear limits on automated monitoring and decision-making practices, including limits and some prohibitions on the collection and use of personal data.
Require transparency of automated monitoring and decision-making, and guarantee worker involvement in implementation, evaluation and impact assessment.
Regulation must also include requirements for transparency of automated monitoring and decision- making, and guarantees of worker involvement in the implementation, evaluation and impact assessment.
Have government regulatory-gap analysis for the tripartite AI Employment and Workplaces Forum draw on evidence from government inquiries and research.
At the time of writing, further analysis of gaps in regulation is being undertaken by the Government to inform the deliberations the tripartite AI Employment and Workplaces Forum.44 This analysis should draw on the significant body of evidence from government inquiries, and research
Underpin regulation and guidance for AI with ethics principles and standards.
Regulation and guidance must be underpinned by ethics principles and standards.
Require data minimisation, limit data collection and use to specific, necessary and justifiable work-related purposes, and give workers and unions notice before AM is used.
Practices should adhere to data minimisation principles limiting collection and processing to the bare minimum necessary to accomplish a given purpose. In addition, data collection and processing must be limited to collection and use for specific, necessary and justifiable purposes connected to the performance of the work. Workers and unions should be given notice before AM is used.
Draw 'red lines' prohibiting algorithmic management practices with a high risk to workers' human rights, dignity, safety and security.
‘Red lines’ should be drawn, prohibiting practices with a high risk to workers’ human rights, dignity, safety and security, such as: o monitoring and collection, processing and use in decision-making of data collected via highly intrusive forms of surveillance, including surveillance of workers outside of work (including when on breaks), and monitoring of private conversations o the collection of data relating to workers’ emotional or psychological state o the manipulation of workers emotional or psychological state o the processing of data to infer union activity and the use of AI to predict or dissuade workers from exercising their legal rights53 o the sale of personal data to third parties.
Prohibit monitoring, collection, processing and decision-making use of highly intrusive surveillance data, including surveillance outside work, during breaks and of private conversations.
‘Red lines’ should be drawn, prohibiting practices with a high risk to workers’ human rights, dignity, safety and security, such as: o monitoring and collection, processing and use in decision-making of data collected via highly intrusive forms of surveillance, including surveillance of workers outside of work (including when on breaks), and monitoring of private conversations o the collection of data relating to workers’ emotional or psychological state o the manipulation of workers emotional or psychological state o the processing of data to infer union activity and the use of AI to predict or dissuade workers from exercising their legal rights53 o the sale of personal data to third parties.
Prohibit the collection of data relating to workers' emotional or psychological state.
‘Red lines’ should be drawn, prohibiting practices with a high risk to workers’ human rights, dignity, safety and security, such as: o monitoring and collection, processing and use in decision-making of data collected via highly intrusive forms of surveillance, including surveillance of workers outside of work (including when on breaks), and monitoring of private conversations o the collection of data relating to workers’ emotional or psychological state o the manipulation of workers emotional or psychological state o the processing of data to infer union activity and the use of AI to predict or dissuade workers from exercising their legal rights53 o the sale of personal data to third parties.
Prohibit the manipulation of workers' emotional or psychological state.
‘Red lines’ should be drawn, prohibiting practices with a high risk to workers’ human rights, dignity, safety and security, such as: o monitoring and collection, processing and use in decision-making of data collected via highly intrusive forms of surveillance, including surveillance of workers outside of work (including when on breaks), and monitoring of private conversations o the collection of data relating to workers’ emotional or psychological state o the manipulation of workers emotional or psychological state o the processing of data to infer union activity and the use of AI to predict or dissuade workers from exercising their legal rights53 o the sale of personal data to third parties.
Prohibit processing data to infer union activity and using AI to predict or dissuade workers from exercising their legal rights.
‘Red lines’ should be drawn, prohibiting practices with a high risk to workers’ human rights, dignity, safety and security, such as: o monitoring and collection, processing and use in decision-making of data collected via highly intrusive forms of surveillance, including surveillance of workers outside of work (including when on breaks), and monitoring of private conversations o the collection of data relating to workers’ emotional or psychological state o the manipulation of workers emotional or psychological state o the processing of data to infer union activity and the use of AI to predict or dissuade workers from exercising their legal rights53 o the sale of personal data to third parties.
Prohibit the sale of workers' personal data to third parties.
‘Red lines’ should be drawn, prohibiting practices with a high risk to workers’ human rights, dignity, safety and security, such as: o monitoring and collection, processing and use in decision-making of data collected via highly intrusive forms of surveillance, including surveillance of workers outside of work (including when on breaks), and monitoring of private conversations o the collection of data relating to workers’ emotional or psychological state o the manipulation of workers emotional or psychological state o the processing of data to infer union activity and the use of AI to predict or dissuade workers from exercising their legal rights53 o the sale of personal data to third parties.
Require employers to inform workers and unions about the deployment and function of algorithmic systems, including monitoring, and ensure automated decisions affecting workers are reasonable, fair and fully justified.
Employers must inform workers and unions about the deployment and function of algorithmic systems and tools, including monitoring systems. Automated decisions that affect workers must be reasonable, fair and fully justified.
Require meaningful, direct human intervention in critical employment-related decisions.
Critical employment-related decisions must not be made without meaningful and direct human intervention.
Give workers the right to challenge decisions that affect them.
Workers must have the right to challenge decisions that affect them.
Mandate consultative mechanisms giving workers a voice, including consultation before deployment of AM systems and worker and union involvement in ongoing evaluation and impact assessments.
Workers must have a voice in the governance and practice of AM. This should be guaranteed through mandatory requirements for the establishment of consultative mechanisms that include consultation prior to deployment of AM systems and tools. Workers and union must be involved in mandatory, ongoing evaluation and impact assessments.
Establish consultative mechanisms that provide for non-unionised workers and workplaces.
Consultative mechanisms must be established that provide for non-unionised workers and workplaces.
Provide workers and unions with adequate information and resources, supported by digital literacy education initiatives, to exercise their rights effectively.
Workers and their unions must be provided with adequate information and resources and supported with digital literacy education initiatives to enable them to exercise their rights effectively.
Prioritise amending existing regulation and introducing new rules where needed, with stronger protections and worker involvement to prevent unacceptable costs and share benefits.
As a priority, the Australian government should review and amend existing regulation as necessary to build in these requirements and introduce new regulation where existing regulation is not fit for purpose. This must include stronger protections for workers grounded in requirements for increased worker involvement in shaping the application of AI at work to ensure it does not come with unacceptable costs, and that any benefits are shared.