Discover how AI agents are changing the rules of the workplace – from automated roles and new governance models to successful implementation. Learn more now about practical examples, HR responsibilities and the new dynamics in the workplace.
Two years after the ‘ChatGPT moment,’ AI hype is a staple of every technology conference – but in many companies, the big breakthrough has yet to come. The tools are there, expectations are high, but implementation is stalled. Language models have democratised AI: no computer science degree or data science background is necessary. Nevertheless, real transformation often fails to materialise. Only with the advent of AI agents is a new picture beginning to emerge – one that changes the rules of the game.
‘Agents’ – a term currently blowing through every tech pitch – is becoming a symbol of the next evolutionary step. They no longer merely respond to inputs. They think for themselves, pursue goals, orchestrate processes and make decisions.
What used to be science fiction is now organisational reality. Not as a tool, but as an actor.
The first wave – chatbots, co-pilots, assistants – thrived on dialogue. Users ask questions, AI responds. Useful, efficient. But agents are in a different league. Three things set them apart: they understand their environment, analyse it independently and derive actions from it. In digital terms, this means they write emails, schedule interviews, formulate assessments – and learn as they go.
The exciting thing is the scale. Instead of sitting opposite an AI, in future dozens will work in parallel – specialists for individual tasks, coordinated by higher-level ‘master agents’. This results in structures reminiscent of human teams. Only faster, more data-driven and permanently alert. The boundary between machine and organisation is becoming blurred.
This reality has already arrived in software development. There, agents have long been orchestrating code flows, monitoring quality and delivering scalable results. Instead of individual work steps, they optimise entire value chains. The efficiency gains are not marginal. They are changing what we understand by work.
When agents take on tasks, the architecture of jobs also changes. The classic job – a fixed set of tasks – is being shaken up. Roles are becoming more important. Tasks are being recombined, skills are being weighted differently, and technology is becoming part of the team.
Agents should not just function. They must be explainable, comprehensible and controllable. This is the only way to create the basis for acceptance – and thus for effectiveness.
Josh Bersin's model of the agile organisation provides the appropriate framework: jobs are broken down into roles, roles are based on skills – and these can be supplemented or taken over by agents. While task-based agents automate individual tasks, role-based agents take on more complex roles with a clear goal in mind. They are more than tools. They are co-players.
An example: the recruitment agent. They screen applicant data, prioritise candidates, communicate with managers and initiate succession processes. All of this is done proactively, scalably and, crucially, integrated into existing systems. The individual process becomes a system process. People remain in the lead. But they no longer use tools – they work in tandem with a digital colleague.
The more tasks agents take on, the more urgent the question becomes: Who is responsible? When machines act, new principles for organisation, ethics and security are needed. HR plays a central role in this.
Agents must be introduced like new employees: they need access, roles and clear frameworks. They must be located – in the structure, in the process, in the responsibility. Workday is developing an ‘Agent System of Record’ for this purpose – a kind of operating system for digital workforces. It brings order, structure and governance to a system that would otherwise be almost impossible to control.
The principle behind it: trust through transparency. Agents should not just function. They must be explainable, traceable and controllable. This is the only way to create the basis for acceptance – and thus for effectiveness.
Every technological revolution begins with a crisis of trust. AI is no exception. Anyone who delegates decisions to systems wants to know how they are made. What has been trained? What data is used? Where are the biases? And who retains control?
Workday is committed to maximum transparency: with AI fact sheets for every function, granular control over data usage, and human final decision-making on critical issues. The systems deliver – but humans make the decisions. Always.
But trust alone is not enough. It needs to be applied. And that's where the real challenge begins.
Technology alone does not transform anything. It is the people who use it – or choose not to. This is exactly where the AI Business School comes in. Its diagnosis: technological development is racing ahead, but many workforces are putting the brakes on. The result: the ‘AI adoption gap’.
This gap can only be closed if two things happen simultaneously: technological introduction and targeted empowerment. Success factor: critical mass. When 15 to 20 percent of the workforce is truly AI-competent, a dynamic emerges that multiplies on its own. Three things are crucial: Training must be role-based, because an accountant needs different tools than a marketer. It must be personalised, depending on prior experience, position and learning curve.
The questions are no longer: Is AI coming? But rather: Who is designing it? Who is using it? Who is growing with it?
And it must be interactive, because learning comes from application, not PowerPoint.
The questions are no longer: Is AI coming? But rather: Who is designing it? Who is using it? Who is growing with it?
At the same time, the job market is changing. Recruiters today compare not only CVs, but also tool skills. Those who work with agents are more productive. Those who can lead agents become talent magnets. The future belongs to those who master both: expertise and AI skills.
Agents are not a trend. They are infrastructure. They are not going to disappear again. And they cannot be ignored. If you want to shape them, you need to act now – clearly, structurally and courageously. Because one thing is certain: the future will not be analogue.
Next step: from theory to practice – with the AI Masterclass
If you want to understand and use agents, you need more than just curiosity. You need knowledge that works. Starting in August 2025, the AI Business School and Workday will offer an exclusive AI Masterclass for HR professionals and executives – in four live virtual sessions with best practice examples, the latest technologies and specific areas of application.
Participation is free of charge for invited Workday or AI Business School customers. After the masterclass, participants will receive two months' access to the AI Business School's globally unique ‘AI Adoption Platform,’ including certificate options.
Register now and play an active role in shaping change.
