How Will Artificial Intelligence and Machine Learning Impact OHS?

Significant technological disruption is occurring across numerous sectors and industries, and this is drastically changing the nature of work and presenting new challenges for work health and safety, according to Macquarie University.

While automation has been ubiquitous in the completion of physical, repetitive workplace tasks for some time (such as product assembly, for example), recent advancements in artificial intelligence (AI) and machine learning (ML) are now bearing major impacts on primarily cognitive tasks, said Dr Ben Morrison, Senior Lecturer in Organisational Psychology at Macquarie University.

AI systems have the potential to benefit the health and safety of workers, such as reducing the cognitive complexity present in many work roles, and conversely, the mental burden placed on workers, said Morrison, who was speaking ahead of SAFEfest for 2021: Health & Safety in a Changing World, which will be held online and in-person from 24- 26 August 2021.

“However, we can also expect a number of arising hazards from their use,” he said.

“For instance, increased AI presence in the workplace may result in reduced contact with other workers and peers and decreased social support, accelerated work pace, and/or increased pressure to incorporate technology-related skills and knowledge alongside workers’ existing expertise.”

Concerns among workers that advances in AI represent a threat to their job security can also adversely impact their wellbeing, explained Morrison, whose research aims to better understand workers’ interactions with intelligent decision support systems and how to ‘design out’ the reasons underlying wrongful rejection, which, in safety-critical situations, may save lives.

“These factors are all expected to introduce significant strain on many workers leading to the increased potential for mental health issues, particularly anxiety and depression,” said Morrison, who observed a number of WHS risks in particular stem from how users choose to interact with AI-driven systems.

“For instance, in relation to overreliance, we see risks of worker boredom, which may impact workers’ ongoing satisfaction and wellbeing,” he said.

“Relatedly, the longer-term impact of ‘deskilling’ may introduce risks to safety in the event that the human worker is required to make unaided judgements on short notice.

“Conversely, in instances of under-reliance, we may see an increased potential for errors in operator judgement, which may pose a risk to workers’ and others’ safety in many work settings.”

Critically, Morrison said there exists a lack of consideration of the WHS effects on workers using AI-driven tools in the empirical evidence base.

“The fact that role- and task-specific factors are expected to play a significant role in understanding the implications to workers’ WHS underlines an urgent and developing problem for WHS professionals,” said Morrison.

Morrison will be speaking at SAFEfest for 2021: Health & Safety in a Changing World. The Victorian, New South Wales, Tasmania and ACT branches of the AIHS will convene a hybrid event with 24-26 August 2021 with virtual sessions, as well as state-based face-to-face sessions in NSW, ACT, VIC and TAS. For more information email events@aihs.org.au or visit the event website.

Article originally published by the Australian Institute of Health and Safety.