Accountability in the age of AI

AI has been steadily changing our professional lives — from the nature of our work to people management practices in modern workplaces too. This shift is altering employee roles significantly and influencing organizational cultures, at large. In an earlier article, I had outlined the growing need for organizations to embrace AI at work. In this one, let’s explore how AI is changing the way we approach work and how we may need to adjust our approach.

AI: Changing the way we think about work

For years, employees were responsible for completing tasks. Now, as AI takes over routine duties, the quantum of accountability is growing. Employees are no longer just executors of tasks; they are accountable for the outcomes. This shift is significant. As Janusz Marcinkowski, Digital Workplace Innovation Consultant at Atos, puts it, "We will become employees accountable for outcomes, not just responsible for tasks."

Traditionally, accountability was the domain of leaders and managers. They set goals and expectations and were accountable for results, while employees focused on performing specific tasks. Today, employees at every level can delegate some of their tasks to AI, but they remain accountable for the results. Marcinkowski adds, "Employees may choose how to complete a task — whether they want to do it themselves or delegate it to AI — but they remain accountable. This marks a shift from simply following instructions to taking ownership of results."

Many employees still confuse responsibility with accountability. They can list their duties but struggle to define the outcomes they are truly answerable for. In the modern workplace, employees choose methods, select tools, and decide what 'good enough' looks like. This is a fundamental change. We no longer wait for instructions — we take responsibility for the effect. Marcinkowski concludes, "My job is to bring R&D initiatives. I don’t wait for someone to tell me what to do — I define my own goals and take full ownership of the results. That’s what being accountable means."

A new model of management: From control to trust

As work changes, leadership must evolve too. AI is redefining what it means to manage. Measuring effort no longer makes sense. Date Reitsema, explains, "Managers would check every detail of what employees were doing. But now the mindset should be: we agree on the outcome — how you get there is up to you."

This marks a deep, fundamental culture shift from control to trust. In many organizations, especially those with hierarchical structures and micromanagement tendencies, giving employees autonomy can be a challenge. Reitsema adds, "Trust is fundamental in changing culture."

But trust doesn’t mean letting go of expectations. It means shifting them. From checking work in progress to understanding how results were achieved, that’s where transparency comes in. Reitsema concludes, "AI also brings a thought process with it. If you ask the right question, it shows you: I got this from here, here, and there. That’s why it should become common practice to include your thought process in the outcome you deliver."

Transparency doesn’t mean micromanagement; it means showing your reasoning. Managers don’t need to see every minute spent, but they do need context: what tools were used, what assumptions were made, and what information was trusted. Transparency isn’t just helpful. It's expected. Showing your process, your choices, and how you use AI isn’t a bonus, it’s part of doing the job well. Because in this new way of working, accountability means owning both the outcome and the path you took to get there.

Skills for the future: Preparing for change

AI is becoming part of everyday work across roles and functions. Every employee now needs a basic knowledge of AI: what tools exist, what they can do, and what their limits are. Bilyana Lyubomirova, Global Head of Career Management at Atos, notes, "If I’m about to recruit somebody now, I think this would be one of the key deciding factors. I would be curious to know how familiar they are with AI tools."

It’s not just about formal skills — exploration matters. Trying different tools and learning by doing are just as valuable as traditional training. Curiosity and a willingness to experiment are in demand more than ever before. Lyubomirova adds, "Even knowing what’s possible, what’s out there, is part of the skill. It's no longer enough to wait for someone to teach you. You need to explore."

At the same time, working with AI requires more than technical know-how. The ability to navigate large amounts of information, judge the quality of sources, and make informed decisions is critical. Lyubomirova concludes, "Working with big data models is very important nowadays. If you used one source of information twenty years ago, now you use 20. You must summarize the information and get the best solution."