Recommendations aim to ensure that the AI ecosystem is genuinely human-centred, principles-driven and accountable.
The All Party Parliamentary Group on the Future of Work has issued a report following its inquiry on the future of work. The inquiry was established in May 2021 in response to growing public concern about AI and surveillance in the workplace. It examined the use and implications of surveillance and other AI technologies used at work; and considered practical policy solutions to meet the challenges and opportunities it found.
Its recommendations are aimed at ensuring that the AI ecosystem is genuinely human-centred, principles-driven and accountable to shape a future of better work. They are centred around a proposal for an Accountability for Algorithms Act.
The key findings
There has been a marked increase in the use of AI technologies in the workplace. Use of algorithmic surveillance, management and monitoring technologies that undertake new advisory functions, as well as traditional ones, has significantly increased during the pandemic. However, the evidence demonstrates that the impacts of AI on work and workers are wide ranging beyond the usual concerns of surveillance or substitution. The inquiry found that the pace, depth and breadth of wider workplace transformation has accelerated. AI technologies are changing the nature of work, who does it and how it is done.
In particular, pervasive monitoring and target setting technologies are associated with pronounced negative impacts on mental and physical wellbeing as workers experience the extreme pressure of constant, real-time micro-management and automated assessment. There is therefore a lack of trust in AI, but the inquiry also found that new regulation could change things. This comes as the UK government has proposed reducing the current protections around AI in the UK GDPR and Data Protection Act 2018 in its data protection consultation.
An Accountability for Algorithms Act
The proposed Act would establish a simple, new corporate and public sector duty to undertake, disclose and act on pre-emptive Algorithmic Impact Assessments (AIA). This duty would apply from the earliest stage of design and deployment of algorithmic systems at work and require rigorous ex ante assessment and ex post facto evaluation of risks and other impacts on work and workers. AIAs would always include a dedicated equality impact assessment.
Updating digital protection
The Act would improve essential protection for workers in response to specific gaps in protection from adverse impacts of powerful but invisible algorithmic systems. These would include an easy-to access right for a full explanation of purpose, outcomes and significant impacts of algorithmic systems at work, a summary AIA and means for redress. A right to be ‘involved’ in shaping the design and use of algorithmic systems at work would be introduced to help better manage impacts on work and workers and to safeguard the social license and democratic governance of these systems. These new rights would be set out in a dedicated schedule to the AAA called “Worker Rights for the age of AI”.
Enabling a partnership approach
To apply the principle of collaboration in the 2021 Digital Regulation Plan and recognise the collective dimension of data processing, the report says that some additional collective rights are needed for unions and specialist third sector organisations to exercise new duties on members’ or other groups’ behalf.
Enforcement in practice
The report says that the joint Digital Regulation Cooperation Forum should be expanded with new powers to create certification schemes, suspend use or impose terms and issue cross-cutting statutory guidance, to supplement the work of individual regulators and sector-specific standards. The forum should be equipped and funded to run regulatory sandboxes to pilot new approaches to actively promote equality, as well as to rigorously enforce existing and new obligations.
Supporting human-centred AI
The principles of Good Work should be recognised as fundamental values, incorporating fundamental rights and freedoms under national and international law, to guide development and application of a human-centred AI Strategy. This would aim to ensure that the AI Strategy works to serve the public interest in vision and practice, and that its remit extends to consider the automation of work. In parallel to the AI Strategy, the Cabinet Office should initiate a Work 5.0 Strategy to address the challenges and opportunities of automation as a result of AI and other modern technologies and work towards a human-centred transformation of work across the UK.