The fourth industrial revolution
The spread of the use of artificial intelligence across all areas of work, including the management of people, continues apace in what has been dubbed “the fourth industrial revolution”.
AI is increasingly being used for all manner of people management functions, for example, to recruit, to train, to line manage, to structure teams and even to make selections for redundancy.
Coronavirus and remote working
Adding to this trend, the advent of mass remote working during the coronavirus crisis has caused employers to rapidly adopt more new technologies and management practices, turbo-boosting the workplace revolution.
According to recent research by the LSE’s Centre for Economic Performance, more than 60% of firms have done so since the start of the pandemic, with 90% saying they expect to keep the changes in place.
Tech for workers
There is great scope for these new technologies to be used to improve people’s working lives, increase fairness and equality at work, and provide evidence and support for campaigns to improve workers’ conditions of employment.
For example, AI could be used to ensure working time or equal pay laws are respected, or to process worker data to evidence trade union campaigns.
But our research suggests that, at the moment, the way AI is being used is often contrary to workers’ interests.
AI and unfairness
Not enough attention is being given to the worker experience of these technologies, and the impact on their well-being, their privacy, their right not to be discriminated against, and their rights over their personal data.
Workers reported to us a variety of problems, including difficulties challenging decisions that had been made about them by AI. There was unfairness in decisions relating to targets, performance management and disciplinary decisions.
Not only this, but there is a strong body of evidence to suggest that discriminatory algorithms are at play. For example, the use of facial recognition and speech recognition technology in video interviewing may discriminate against disabled applicants.
Unfairness of outcome and lack of consultation and transparency appears to have created a lack of trust in AI technologies. Indeed, only 28% of workers are comfortable with technology being used to make decisions about people at work (Britain Thinks Polling commissioned by the TUC).
We also heard from workers that the way AI is used to manage and monitor them has a negative impact on their mental and physical health, with one describing their workplace as “increasingly robotic, alienating, monotonous and lonely” and another explaining that “going to work is not enjoyable anymore as you are scrutinized and watched over constantly”.
Workers also reported problems with equal access to the use of AI-powered tools, for example, for training purposes, with blind colleagues sometimes excluded.
There are big commercial gains to be made for employers who can make efficiencies in their human resource functions by using AI- powered tools and platforms, as well as for the tech companies that produce the software packages.
As a result, commercial and employer interests are well-represented when AI is developed and used to manage people at work, but the worker voice is not. Only 31% of workers agree that staff at their workplace are consulted when any new forms of technology are introduced (Britain Thinks polling commissioned by the TUC).
This imbalance can only be redressed with:
a strong collective voice to bring attention to the worker experience of being managed by AI,
access to understandable information about the technologies being used and how they work,
fairness and equality in the application of AI, and
more worker control over their data.
As with any tricky problem, collaboration and communication is key, so we call for technologists, workers, trade unions, employers and government to work together to ensure a future where AI works for everyone.
Want to hear about our latest news and blogs?
Sign up now to get it straight to your inbox
To access the admin area, you will need to setup two-factor authentication (TFA).