Originally published in Svenska Dagbladet by Björn Jeffery, September 16, 2025
Swedish Sana is being acquired for 10 billion kronor by American Workday. The timing of the deal could prove historically good — is Sana founder Joel Hellermark seeing something no one else sees?
The question every board is currently asking its management team is: what is your AI strategy? And when the answer sounds a little too hollow, it is time to open the wallet. Generously. The fact that American software company Workday is now acquiring Swedish Sana for 1.1 billion dollars — around 10 billion kronor — is one example of that.
The deal gives both sides what they need. Sana gains access to Workday’s customers and the 75 million users it has there. Workday gets an alibi that says it is part of the AI game. But the central element of the deal is the timing. CEO Joel Hellermark has a reputation for having a deep understanding of the AI market. Could what he has seen have been behind the decision to sell?
Despite the incredible hype, it is worth reminding ourselves that AI implementations among larger companies have not yet delivered the big economic numbers. Some areas, such as programming assistance, have shown promising results. AI writing better code is the category most commonly cited by those who have tested it as genuinely useful.
Even so, AI coding tools are areas where AI can be useful but they are not where the biggest economic gains are to be found. Among the futurists, increased economic productivity has been promised. We are talking about efficiencies: employees performing tasks with superpowers, and perhaps fewer employees altogether. For that to happen, large-scale adoption by companies is required.
Sana is operating in precisely that market. The company now offers services to streamline how companies work with internal documents and systems. Keep the IT structure you have, but add AI on top of everything. It is an appealing pitch for many company leaders who may have been worried about the cost this AI revolution might entail.
At the same time, the promise of these efficiencies has not always been delivered. A new survey from the US Census Bureau shows a distinct trend break in how large companies with more than 250 employees have implemented AI. The figure has been falling over the past two months — despite the incredible tailwind the AI trend has been enjoying.
Another study from MIT found that only 5 percent of all new generative AI initiatives were profitable. Ninety-five percent were losing money. That investment in new and immature technology does not pay off immediately is perhaps not surprising. But the figure could dampen enthusiasm among companies that have tested it. It is easy — perhaps wrongly — to write off a technology too early. But it nonetheless becomes a problem for those trying to sell it.
This brings us back to Joel Hellermark and the timing of the deal. Who says no to 10 billion kronor, one might reasonably ask? But someone who has been as deeply embedded in the AI world as Hellermark — and who has seen the momentum his industry has had — could definitely have done so. The former CTO of OpenAI, Mira Murati, had her newly founded company valued at 110 billion kronor straight away. Mark Zuckerberg at Meta is paying billions of kronor to individual employees.
So why is Hellermark selling his life’s work now? Perhaps because he sees something others do not yet see. That it will take longer to get the world AI-adapted than many believe. And that a potential rough reckoning for the entire industry could be a cold shower for everyone participating in it.
In a scenario like that — you get off. You take your billions and dock yourself alongside a giant, listed, and slightly dry company like Workday. The timing for Hellermark and Sana could prove quietly historic.
