Google Pledges Up to $40 Billion to Anthropic in AI Race

Google’s pledge to invest up to $40 billion in Anthropic showed how expensive the foundation-model race has become, and why the fight is shifting from flashy demos to the plumbing underneath enterprise software. The deal includes $10 billion upfront and as much as $30 billion more tied to performance milestones, a structure that makes compute and model access look less like background costs and more like strategic assets.

The scale matters because Google and Anthropic are not just partners. They are also rivals in the global AI race, and the transaction deepens a relationship that has long been tied to Google Cloud and TPU infrastructure. Amazon has already expanded its own commitment to Anthropic to as much as $25 billion, underscoring that the leading model labs are increasingly backed by multiple hyperscalers at once. For enterprise buyers, that is a warning sign and a promise at the same time: the best AI features may arrive faster, but the stack behind them could become more expensive and more fragmented.

For monday.com, the immediate lesson is practical. If model pricing, inference costs, or partner access changes quickly, product managers and finance teams have to think about packaging and usage limits earlier in the roadmap. Engineers need abstraction layers that can route or replace models without forcing a rewrite of every workflow. Sales teams need a clearer answer for customers who now know that enterprise AI features sit on top of a highly contested supplier market, not a stable utility.

That pressure lands at a moment when monday.com is pushing hard to define itself as an AI work platform, not just a work-management tool. The company says more than 250,000 customers worldwide use its platform. In its 2025 results, monday.com said revenue grew 27% for the year, fourth-quarter revenue reached $333.9 million, customers with more than $50,000 in annual recurring revenue represented 41% of total ARR, and it recorded net adds of customers with more than $100,000 in ARR. In March 2026, monday.com launched Agentalent.ai, a hiring platform for enterprise AI agents, and earlier that month it said external AI agents would be able to access the platform and operate alongside human users.

That is the real signal inside the Google-Anthropic deal. Enterprise software buyers are moving beyond generic AI checkboxes and asking which models power each feature, whether those models can be replaced, and how vendor concentration affects price and reliability. For monday.com, the next test is whether its modular, agent-oriented pitch can keep pace with a market where the model layer is becoming both more powerful and more costly.