CEO Sebastian Siemiatkowski and Zoom’s Eric Yuan are among the first to experiment with AI versions of themselves for public-facing presentations. At Klarna, an AI-powered avatar of Siemiatkowski handled most of the fintech firm’s recent earnings call — with only subtle giveaways, such as a slightly out-of-sync voice, hinting that it wasn’t the real CEO. The fact became apparent when the avatar disclosed its AI identity.

Soon after, Zoom’s Yuan followed suit. His AI avatar delivered the opening remarks at the company’s earnings call using Zoom Clips, its internal video tool.

“I am proud to be among the first CEOs to use an avatar in an earnings call,” said the digital Yuan. “It is just one example of how Zoom is pushing the boundaries of communication and collaboration. At the same time, we know trust and security are essential… We’ve built strong safeguards to prevent misuse, protect identity, and ensure avatars are used responsibly.”

This move comes on the heels of a Harvard Business Review study published last September, which tested GPT-4o’s performance in a simulated CEO role. The results: the AI outperformed human participants on most metrics — yet it was fired sooner.

The AI failed during simulated market shocks, akin to the unpredictability of the Covid-19 pandemic. Human participants, mostly business students from top global universities, adopted more conservative strategies, avoided risky inventory, and emphasised flexibility and sustainable growth. GPT-4o, on the other hand, pursued aggressive expansion and quick wins — a strategy that fell apart when volatility struck.

The study concluded:“Despite its impressive performance, AI cannot assume the full responsibility of a CEO in markets that serve humans. Instead, it can significantly improve the strategic planning process and help prevent costly mistakes.”

In other words, the chaos and unpredictability of real-world markets — and the need to navigate human relationships — still make the CEO job uniquely human. At least for now.