Agentic AI demands more humanity: the role of human skills in AI adoption

Agentic AI is changing how we work, that is a fact. Where AI used to mainly support, today we see agents that independently carry out tasks, prepare decisions and drive workflows. According to some insights from Microsoft and WorkLab, this evolution is described as a shift towards human–agent teams: AI executes, people determine direction and remain ultimately responsible.

That reality does not call for faster tool adoption, but for a different view on Digital Change.

‍

From learning to use technology to reshaping work

“The technology itself is surprisingly accessible today,” says Graciela Zaera Moraña, team lead Digital Change & Adoption at The Flow. “We work in natural language. The real challenge lies in finding valuable scenarios and embedding them in new working habits.”

Digital transformation is therefore less about tools and more about rethinking work: what do we let run autonomously, where is human judgement crucial and how do we keep control?

Human at the wheel, always

Even when AI can make decisions independently, the human remains ultimately responsible. Assessing AI output, adjusting it and making conscious choices remains essential.

According to Graciela, that responsibility cannot be placed on individual employees. “Organisations must establish clear rules of play: who decides, who checks and where human expertise remains indispensable.”

Without that clarity, uncertainty arises, and with it trust disappears.

The 5 human skills that give Agentic AI direction

As AI becomes more autonomous, value shifts to human skills that cannot be automated. WorkLab names five essential human skills, the 5 C’s:

• Curiosity: keep learning and discovering

• Critical thinking: judging and adjusting

• Creativity: creating value and solutions

• Collaboration: working together with AI and people

• Compassion: taking people along in change
  

Agentic AI therefore calls not for less, but for more humanity. That is why we translate these 5 human skills into concrete Digital Change journeys, in which leaders and teams learn how to consciously steer Agentic AI in their daily work.

But where does it pinch for organisations?

In many companies, critical thinking and collaboration remain underexposed. It is precisely these skills that are crucial when people have to assess AI output and share responsibility.

“Critical thinking today also means: continuing to grow as a domain expert,” says Graciela. “AI supports, but above all strengthens those who keep actively learning themselves and can substantiate choices.”

What it requires of leaders

Effective collaboration with Agentic AI requires trust in technology, in one’s own judgement and in clear organisational frameworks. And that trust starts with leadership.

Agentic AI calls for leaders who give direction, set boundaries and take ownership. Classic change management too often focuses on tool adoption and too little on human judgement and responsibility.

“Leaders must take on this role themselves first. Only then can Agentic AI land sustainably in the organisation.”

Agentic AI, every single day

With us, Agentic AI journeys do not start from features, but from human behaviour. We translate the 5 C’s into leadership guidance, habit formation and context-oriented training within Microsoft 365 and Copilot.

Not training per role, but guidance in realistic work scenarios, so that people not only use AI, but also consciously steer it.

Collaborating with AI is no longer a future vision, but a daily reality. Or as Graciela summarises it: “For every task I ask myself one question: can Copilot help me with this?”