Salesforce Inc. (CRM) said Tuesday it will invest $1 billion in Switzerland over the next five years to accelerate the country’s adoption of agentic AI, according to a Seeking Alpha report.
The pledge comes as Salesforce shares trade near a two-year low, weighed down by investor doubts about whether all of its AI spending will pay off.
Switzerland is not an isolated bet. It is the latest entry in a spending campaign that has quietly become one of the most aggressive corporate AI investment pushes anywhere.
Chair and CEO Marc Benioff announced the commitment during a visit to Geneva ahead of the AI for Good Global Summit, where he will co-chair the inaugural meeting of the new AI for Good Global Commission alongside Rwandan President Paul Kagame and International Telecommunication Union Secretary General Doreen Bogdan-Martin, according to a Salesforce statement.
Geneva was not a random stop. The city hosts the ITU and the World Economic Forum and will host the Global AI Summit in 2027.
Related: Salesforce makes gutsy bet to win AI agent race
Salesforce will fund Swiss workforce expansion and AI training
The money is meant to fund Salesforce’s Swiss workforce, its growing customer and partner base, and local AI skills training, according to a company statement.
Switzerland now joins a growing list of countries receiving similar pledges.
Salesforce committed $1 billion to Italy in June and $2 billion to France that same month, on top of an earlier $3.5 billion five-year commitment there.
Salesforce points to existing traction in Switzerland as proof of concept. Virtual care provider Oviva uses Agentforce to handle more than 300,000 monthly customer messages autonomously, deflecting half of all inquiries without a human involved, the company said.
Bag maker FREITAG built an agent called FRIDA for customer service, and the World Economic Forum deployed an agent named EVA to guide more than 3,000 leaders through its 2026 Davos meeting.
LUDOVIC MARIN / Getty Images
What is agentic AI, and why is Salesforce racing to own it?
Agentic AI differs from the chatbots most people already know. Instead of only answering a question, an agent can complete a multi-step task on its own, such as resolving a support ticket from start to finish.
That distinction matters commercially because it changes what software companies can charge for.
Salesforce has historically sold per-seat licenses to human employees. If agents start doing part of that work, the pricing model shifts toward consumption or outcomes, and Salesforce needs that shift to happen on its own terms.
More AI:
- The new Chinese AI model rattling U.S. tech investors
- Anthropic restores access to Mythos 5 for select organizations
- SoftBank CEO offers stinging critique of Musk’s AI bet
Agentforce annual recurring revenue reached $1.2 billion in the quarter ended April 30, up 205% year over year, according to Salesforce’s first-quarter fiscal 2027 earnings release.
That is the fastest revenue ramp of any product in the company’s history, though it remains a fraction of the company’s record $41.5 billion in total annual revenue for fiscal 2026.
Investors worry autonomous agents will cannibalize seat-based revenue
Despite that growth rate, Salesforce shares have fallen roughly 33% so far this year, dropping from $253.62 to $169.01, according to CNBC, as investors weigh whether agent-driven efficiency will eventually erode the seat-based licensing business that still generates most of the company’s revenue.
The market has learned to discount forecasts and back-half promises. It wants organic proof instead.
That skepticism helps explain the pattern of country-specific pledges. Locking in enterprise and government relationships now gives Salesforce a claim on future AI budgets in markets like Switzerland, Italy, and France, even while near-term monetization stays murky at home.
The pattern echoes how cloud providers once courted governments with local data centers years before the revenue followed.
Salesforce is running the same playbook with agentic AI, spending capital today for customer relationships and policy access it hopes will matter more once agents take on more of the work companies currently pay humans, and Salesforce, to manage.
Whether that bet pays off may hinge on a question Geneva’s AI summit is grappling with this same week: how much of what agents promise to do, they can actually deliver.
Related: Turing cuts Nvidia reliance, taps AMD for 10% of AI training
