Access to tools, not training, is stifling AI value
Europe’s AI ambitions are stalling, but not due to a skills gap. Limited access to tools and outdated work design are holding companies back....
by Michael Watkins Published September 21, 2026 in Artificial Intelligence • 4 min read
In a clip that Klarna CMO David Sandstrom posted to LinkedIn in 2025, CEO Sebastian Siemiatkowski delivers a piece to camera about Klarna’s Q3 results. Or does he? The man on camera looks and talks like Siemiatkowski.
To a casual viewer focusing on the detail rather than the delivery, nothing would seem amiss. But looking closely, the video has a slightly uncanny feel, as if we are looking at Siemiatkowski’s face from two different angles at once.
Further scrutiny gives rise to the suspicion that he is repeating a limited range of facial expressions in sequence.
And there is one more clue that all is not as it seems: the caption that reads: “For the first time, Klarna’s Q3 update is presented by an AI-generated version of Sebastian Siemiatkowski. This is how AI transforms not just what we do, but how we do it.”
Klarna is not alone in putting up an AI avatar of its leader in the public realm. Meta has hatched its own plans for an AI avatar of CEO Mark Zuckerberg, which he is reportedly hoping to use to interact with staff to improve engagement.
And Zuckerberg has taken the idea one step further: he is reported to be building an AI assistant, ushering in a future where AI agents could perform core CEO tasks.
Klarna’s video is grabbing attention. At time of writing, the post had 1,290 likes. And this has been achieved without Siemiatkowski having to spend time recording the video. But not all reactions have been positive.
Comments underneath the post question how much – if any – time was saved by creating a digital doppelgänger of the CEO. Some also express concerns about gimmickry and even video fraud. So, do the positives outweigh the negative fallout? And what would a wave of CEO AI avatars mean for the future business world?
The immediate worry is reputational. To facilitate the creation of their avatar, the CEO must effectively surrender their likeness to an in-house tech team, or for those without in-house capability, an external partner.
For such a delicate and personal operation, strong guardrails are essential. CEOs should be stringent in controlling access to their avatar, with a watertight sign-off process.
They will also want to retain firm oversight of their avatars’ output, as poorly worded statements or dreaded AI hallucinations could tarnish hard-won human reputations. Even if the audience (for example, the company’s employees) know that they are looking at an avatar, they will consider it to represent the views and attitudes of the real person it resembles.
Another, more insidious, consequence of using an AI-generated likeness of a leader is the erosion of belief in the CEO themselves.
Saturating social media with AI-generated avatars of CEOs risks distancing stakeholders and the wider public from the actual human expertise of the CEO; the strategic experience, communications skill and charisma which marks out someone fit for leadership.
Eventually, the workforce will feel distanced and less inspired by their actual human CEO, potentially devaluing the role, first within and eventually outside of the company.
Moreover, using an AI avatar for some interactions and not others creates a two-tier system. Who gets to interact with the CEO and who gets the avatar? Those who can’t get past the AI version may come to resent both it and the individual it represents.
Could we ultimately replace CEOs with AI entirely?
Beyond the reputational consequences, what could AI avatars mean for the CEO? Over half (55%) of surveyed CEOs in one survey say they have felt lonely in the role. Handing off much of their public-facing duties to AI could free up time to spend with their teams.
But equally, delegating people-focused activities could mean the CEO spends even more time alone.
Bringing greater AI influence to bear on the role also makes CEOs more vulnerable to the negative effects of some AI tools. There is already some evidence of skills atrophy in junior workers when employees use large language models (LLMs).
If AI avatars take over the tasks that demand the soft skills essential to the CEO role, for example in interactions with stakeholders or presenting results, the real CEO may suffer both a diminution of those skills and the negative psychological effects of feeling dispensable.
This leads us to the question, could we ultimately replace CEOs with AI entirely? A little unnervingly, one company in China has already answered in the affirmative. What’s more, it has outperformed the market in the six months since making the “appointment.”
While the world is watching a digital version of their faces, CEOs may need to watch their backs.
Professor of Leadership and Organizational Change
Michael D Watkins is Professor of Leadership and Organizational Change at IMD, and author of The First 90 Days, Master Your Next Move, Predictable Surprises, and 12 other books on leadership and negotiation. His book, The Six Disciplines of Strategic Thinking, explores how executives can learn to think strategically and lead their organizations into the future. A Thinkers 50-ranked management influencer and recognized expert in his field, his work features in HBR Guides and HBR’s 10 Must Reads on leadership, teams, strategic initiatives, and new managers. Over the past 20 years, he has used his First 90 Days® methodology to help leaders make successful transitions, both in his teaching at IMD, INSEAD, and Harvard Business School, where he gained his PhD in decision sciences, as well as through his private consultancy practice Genesis Advisers. At IMD, he directs the First 90 Days open program for leaders taking on challenging new roles and co-directs the Transition to Business Leadership (TBL) executive program for future enterprise leaders, as well as the Program for Executive Development.
September 18, 2026 • by Mauro Macchi, Tim Good, Dominic King, Mamta Kapur in Artificial Intelligence
Europe’s AI ambitions are stalling, but not due to a skills gap. Limited access to tools and outdated work design are holding companies back....
September 16, 2026 • by Michael R. Wade, Konstantinos Trantopoulos in Artificial Intelligence
The AI Safety Clock enters the critical risk zone as autonomous agents gain authority, military AI expands, and AGI-related risks accelerate....
September 15, 2026 • by Michael R. Wade, Howard H. Yu, Amit Joshi, Naomi Haefner, Faisal Hoque, José Parra Moyano, Mark J. Greeven in Artificial Intelligence
Amid warnings from industry insiders that AI will soon outpace human control, IMD's experts break down where business leaders should be focusing their attention....
September 7, 2026 • by Faisal Hoque, Paul Scade , Pranay Sanklecha in Artificial Intelligence
Record CEO turnover, fractured geopolitics, and an AI transition that most companies are failing to profit from require a fundamental shift in the ways we lead and govern. Here is how to...
Explore first person business intelligence from top minds curated for a global executive audience