AI must be used to benefit workers – not as a means to downgrade jobs and pay

We are at a crucial moment in what has become widely known as the fourth industrial revolution. Artificial intelligence is already being used to carry out many human tasks and the pace of change has been turbo-charged by the coronavirus pandemic. AI-powered tools are increasingly being used to manage people who are working remotely. But few of us realise the extent to which AI is being used by employers to make important decisions about people at work. We need to seize this moment to make sure that AI is used to benefit workers – not simply as another means to downgrade jobs and pay.

More and more, AI is making life-changing decisions, like who gets a job, who gets paid what salary and who gets made redundant. And when these decisions are unfair, discriminatory or unsafe, this has huge consequences for people’s lives. But workers tell us that they are often unaware of AI being used in this way, that it’s difficult to challenge decisions and how they feel constantly scrutinised. Yet AI technology is being rolled out at work too fast for employment law to catch up. We believe society and policymakers now have a choice to make about whether we put human values at the heart of this technological revolution, or whether we allow commercial interests alone to drive developments. What we choose to do now matters to all of us.

What should we do? The TUC aims to seize this moment in the AI revolution to put human interests at the heart of technological developments, and to ensure that workers benefit from the opportunities offered by AI. Today, we have published a short manifesto that sets out the values employers, technologists and political parties should adopt to ensure that technology at work benefits all. Alongside these values, we make a series of proposals, including reform to the law, based on recommendations made by leading employment lawyers Robin Allen QC and Dee Masters and TUC research.

The importance of collective bargaining is a key principle that underpins our manifesto. Some of our proposals are non-negotiable red lines that we believe are critical to ensuring dignity at work: AI should only be used to make important decisions about people if these decisions are easy to explain and understand; no unlawful discriminatory decisions should be made using technology; everyone at work should have a say in deciding whether AI is introduced to make important decisions about people at work; it should be clear when AI systems are being used at work; and there should be enough information available to people at work and job applicants to satisfy them that the technology is being used fairly.

To secure our red lines, we propose a series of amendments to data protection, employment and equality legislation, as well as better guidance on how to avoid AI-based discrimination, and a statutory duty to consult trade unions. But we propose increased regulation to target high-risk applications only, not harmless uses of AI at work. We do not want to inhibit innovation, but we do want to make sure that innovation benefits everyone at work. And in order to reassert human agency in the face of technological control, we propose a new right to human review of decisions made using AI, as well as a right to in-person engagement and a right to disconnect from digital devices.

The TUC is also calling on workers and trade unions to harness the power of data and AI, proposing a ground-breaking new right of “data reciprocity” whereby workers could collect and combine workplace data. This could help identify patterns in who isn’t receiving their rights and why that could assist union campaigning for better pay and benefits at work. And we highlight ways in which trade unions could help workers achieve data equality, including adopting a formal data gathering role, engaging data scientists and developing AI-powered tools.

Our vision for the future world of work is one where technology is for the benefit of everyone, and humanity is at the heart of all development and application of artificial intelligence at work. You can join us in making this vision a reality by signing up to the values and proposals in our manifesto.