Work is no longer done only by people. Yet most org charts still assume it is.

Across industries, leaders are experimenting with AI copilots, autonomous agents and workflow automation. Teams are using agentic AI to write code, generate marketing campaigns, analyse contracts and even triage customer issues. The productivity potential is clear.

Given the magnitude and impact of this technological shift, ZRG is embarking on a series of conversations with senior leaders about how accountability and ownership evolve as humans and machines increasingly share the work.

Policy experts and business strategists alike are beginning to focus on a fundamental question: where does accountability sit when AI acts?

In many organisations, the conversation still revolves around ownership of the technology. Is this an IT initiative? A data initiative? An HR transformation effort? Sometimes the debate centres on who controls the AI budget or the tools.

But the reality is different.

Agentic AI creates a new accountability gap.

As ways of working are redefined and “work” becomes the output of both human and algorithm, accountability for that output becomes less clear. It sits in the spaces between traditional functional boundaries. The leaders who design work, the leaders who deploy technology, and the leaders responsible for business outcomes are now collectively shaping how work actually gets done.

So, when agency is no longer implicitly linked with human ethics and decision-making norms, when resources are not always human, and when agentic AI is democratised rather than centrally governed, who is accountable?

In many companies, accountability still follows legacy lines: Technology teams manage the platforms; HR manages the workforce; Business leaders own the commercial outcomes.

But digital labour sits across all three. So where does the buck stop? And with whom?

But digital labour sits across all three.

Technology leaders control how systems operate. People leaders shape how humans work alongside them. Business leaders ultimately carry the results.

The reality is that accountability is becoming shared by design.

The real shift is not technology adoption. It is work redesign.

The first insight leaders are confronting is that AI adoption quickly becomes a question of operating model design.

Consider software development. Engineers now routinely use AI coding assistants to generate large portions of code. The productivity gains can be significant. But the work itself changes. Developers spend less time writing code and more time reviewing, refining and orchestrating outputs.

Who owns the quality of that output? Who ensures teams have the skills required for this new role construct?

The CIO may oversee the tools. The engineering leader manages the team. But the work has fundamentally changed, and the accountability structure often has not.

The same pattern is appearing across functions. Marketing teams generate content with generative AI. Legal teams use AI to review contracts. Finance teams automate analysis and forecasting.

While some organisations are still adding technology to existing roles, many are moving towards a fundamental redesign of how work is performed. Yet the leadership structures responsible for that redesign are often unclear.

Technology ownership alone will not solve the problem.

Some organisations attempt to address this by assigning ownership of AI to a single functional leader. Often this sits with the CIO or a newly created AI leadership role.

That clarity helps with infrastructure and governance. But it does not fully address the work design challenge.

Technology leaders can design and deploy the tools, but they cannot unilaterally redesign how thousands of employees and machines collaborate across the enterprise.

The same is true for HR. Workforce strategy matters deeply, but HR leaders do not control the enterprise technology environment or the operational workflows where AI is deployed.

The reality is that no single function owns the entire problem.

The organisations that move fastest will treat this as a shared leadership challenge.

The most forward-looking companies are beginning to treat AI not as a technology initiative, but as a work redesign challenge that spans the leadership team.

Agentic AI is now integral to operating models and org charts. But accountability for how work gets done remains firmly with the human leaders who design the system around it.

How leadership teams adapt to this new paradigm, and to the blurring of functional boundaries it is already creating, will define organisational success in the years ahead.

‍