The chasm between anticipated AI shake-ups and current practices

More than half of newly surveyed workers anticipate meaningful role changes due to AI, but fewer are actually doing something about it.

New data from Amplitude found that approximately 40 per cent spend less than one hour a week learning or experimenting with AI. But is this the responsibility of the individual, or a company’s duty?

According to Mark Drasutis, head of value for APJ at Amplitude, it needs to be a shared responsibility. He said: “If businesses want to realise the full value of AI, they cannot rely on employees to figure it out on their own. Upskilling needs to be embedded into the organisation through structured training and clear guidance on how AI should be used in day-to-day work.”

However, he also acknowledged that “those who proactively build their skills will be better positioned to adapt and advance their careers”.

This benefit leans more towards individual advantage, leaving a current problem for organisations. The divide corresponds to age and generation, the data revealed, with just 4 per cent of workers between the ages of 55 and 64 trusting AI over their own judgement, compared to almost a third in the 18- to 24-year-old group. Accordingly, 20 per cent of the former group actually use AI, compared to 39 per cent in the latter.

Drasutis said that “a significant proportion of younger workers are upskilling in AI outside of work hours, which points to a gap in formal organisational support”.

And considering the fact that professionals are more likely to be in senior leadership roles, it is clear how the trust gap is affecting businesses.

Organisational AI integration remains patchy – despite higher AI capability, only 13 per cent of professionals under 24, and 9 per cent of those between 25 and 34, claim AI is core to their work. Overall, almost half of respondents reported improvement in their company’s operational AI use – while acknowledging there is more to be done – and less than a quarter said their organisation rarely uses AI at all.

Fragmented approach to AI adoption, Drasutis said, “creates operational and cultural risks”.

“Our research shows that while AI use is relatively widespread, organisational maturity remains low and age-based discrepancies around trust in AI can be leading to uneven productivity gains,” he added.

Without leadership-led frameworks that provide accuracy and quality, he said, businesses risk adoption ceilings and complex security, governance, and compliance burdens.

He also referenced the potential for cultural gaps “between savvy users and sceptics” whereby uneven adoption is creating friction. According to the survey, 18 per cent of respondents cited colleague competitiveness, and 11 per cent witnessed resentment from non-AI users.

On bridging the multitude of gaps, Drasutis said: “Building a single universal program is often the wrong starting point, as needs vary significantly across generations and experience levels. “

Younger employees, he highlighted, usually require guidance around governance and context, whereas more senior employees often respond to confidence and relevance.

“That means low-stakes environments to experiment, reassurance that AI supports rather than replaces their judgement, and clear examples within their own domain,” Drasutis said.

“Our research found that 5 per cent of workers learn about AI through mentorship or peer learning at work. This is a missed opportunity, and one of the lowest-cost interventions available to HR teams.”

An effective way to harness this, he said, is cross-generational, reverse mentoring, whereby junior employees share technical fluency while senior employees translate their strategic expertise. Drasutis said: “Both benefit, and it helps reduce the tension many organisations are seeing by replacing uncertainty with shared purpose and improving collaboration.”

On effective and impactful adoption overall, Drasutis said it begins “by anchoring learning to real work”.

“Programs built around tasks within an employee’s role are more likely to drive meaningful adoption than generic AI literacy modules. From there, learning to be sustained over time,” he said.

Additionally, “organisations are encouraged to measure behaviour change”.

To fully address discrepancies, Drasutis highlighted, “leaders should lean on time-tested change management techniques so that every employee is brought on the journey”.

“Experimentation is not inherently negative, provided organisations can implement the right balance of guardrails, training, and approved toolsets in place to ensure consistency and security,” he said.

Want to see more stories from trusted news sources?
Make HR Leader a preferred news source on Google.