Managers are far more enthusiastic than frontline workers about the potential of generative AI tools to improve their business, according to a new study that unearthed broad concerns that companies aren’t committed enough to training workers and managing AI responsibly.

While 86 per cent of frontline workers say they’ll need training to understand how AI is going to change their jobs, just 14 per cent have actually received such training, Boston Consulting Group’s new AI At Work survey found.

That was far behind the 44 per cent of business leaders in the study – which included 12,898 managers and employees in mostly large businesses across 18 countries – who said they had received training about AI’s potential to change their business.

And while 62 per cent of leaders are optimistic about the potential of AI, just 42 per cent of frontline users expressed the same.

Surging interest in AI has reduced overall concern about the technology compared to the previous survey, which was conducted in 2018: just 30 per cent of respondents expressed concern about the technology now, compared to 40 per cent in the previous survey.

Yet 36 per cent of workers still fear adoption of AI will eliminate their jobs – a concern highlighted by advances such as Adobe’s recent release of a generative AI tool that could marginalise graphic artists by enabling creation of context-sensitive content with a few words.

“We saw a very significant improvement in positive feelings towards AI and, in particular, optimism,” noted study co-author Nicolas de Bellefonds, global leader of AI and software at the group’s tech build and design unit BCG X, which has worked with over 500 companies to understand generative AI in the past 6 months.

“Employees feel that AI will save them time and promote innovation in their roles and daily lives at work,” he continued.

“But there are some deep-seated issues that employees and leaders are feeling – in particular, the fact that there is a very significant need for upskilling of employees, and that companies are not investing enough.”

But is my job safe?

Although the survey results suggested that workers become more comfortable with AI the more they are exposed to it, it also exposed lingering uncertainty about just what an AI-empowered workplace would look like.

“Almost every respondent believes they will need upskilling,” said BCG X global leader for talent and skills Vinciane Beauchene.

“Although employees realise that AI is going to be a revolution – and is going to have a significant qualitative impact on the nature of their jobs – companies are still not ready to undertake what it takes to adapt to this revolution.”

Given AI’s potential to not only empower employees but to potentially replace them, she said, companies need to approach employee training with a clear purpose and vision.

“When we’re talking about upskilling, we’re not just talking about how to use the technology,” she explained.

“We’re really talking about anticipating much more, to understand the impact of the technology on the nature of the job – and, therefore, the skills that will differentiate your performance in your job.”

Australian companies are particularly exposed in this regard, with a recent Udemy analysis showing that Australian workers are significantly behind international peers and, as Udemy senior vice president of supply strategy Scott Rogers warned, “yet to embrace ChatGPT as a skill”.

Individual businesses will need to formalise their expectations around AI and develop clear transition and education programs to help employees move into the AI-driven future – but “that is going to be a stretch for most companies,” Beauchene said.

“Suddenly, it’s about anticipating the evolution of jobs, better understanding the skills that are currently mastered by your employees, and being much faster when it comes to producing upskilling content.”

As well as delivering ongoing AI training that mirrors the technology’s rapid growth, BCG warned that corporate AI strategies will also need to explicitly address the need for ‘guardrails’ around AI’s use and application – with 79 per cent of ANZ respondents agreeing that AI-specific regulations are necessary.

Here, too, there was a huge perception gap: while 68 per cent of business leaders said they feel confident about their organisation’s responsible use of AI, just 29 per cent of frontline employees felt the same.

Responsible use of AI is “paramount,” said de Bellefonds, noting that BCG is launching a Center for Responsible Generative AI to help its clients address the issue.

“Employees want reassurance that their organisations are approaching AI and generative AI ethically, and leaders want to be in a position to help frame emerging AI regulations.”