According to Mary Towers, an employment rights policy officer at the Trades Union Congress (TUC), research into the uses of AI in the workplace is showing that using technology to recruit and manage workers, replacing the functions of a human manager, can negatively affect both the physical and mental health of workers.

Speaking on Tuesday [21 January 2025] to the All-Party Parliamentary Group on Modernising Employment, which is looking into digital hiring and employment practices, Towers cited TUC research in calling for employers to work hand-in-hand with workers to explore threats posed by algorithms and unregulated AI systems.

“Workers have described to us the dehumanising and sometimes isolating effects of AI-directed work, unsustainable work intensity, the stress of being constantly monitored and assessed,” Towers said.

“But as well as stress and work intensification, it’s also now widely accepted that the use of AI at work can entrench inequality and bias with the outcomes merely replicating baked-in bias from the data that’s been used to train AI,” she said. “The phrase you often hear is ‘garbage in, garbage out’.”

The UNITE union has recently called for an end to discrimination by algorithm and recognising the threat of unregulated AI systems, further embedding workplace discrimination, she noted to the audience.

“Our research also identified a worrying lack of consultation of workers and difficulties understanding how decisions are being made about workers by AI with a lack of transparency about this,” she said.

TUC polling showed that about 70% of workers think that employers should consult their staff before introducing new technologies at work and about 70% also “oppose AI being used in really high-risk decision, such as whether or not someone gets a job and whether or not their employment is terminated”, Towers said.

She added: “It’s really important to make it absolutely clear that we are not anti-innovation – but [support] fair innovation that works for everyone.”

She said the TUC, alongside other organisations, were proposing an AI bill that at its core would “ensure that only safe AI systems make it into the workplace, and that workers and unions are fully consulted, informed and involved”. This would include a risk assessment of AI decision-making, as well as a consultation process that is built into the risk assessment.

For more information about the TUC bill, click here.

• Comment below on this story. Or let us know what you think by emailing us at [email protected] or tweet us to tell us your thoughts or share this story with a friend.