Merck Signs Up to $1 Billion Deal With Google Cloud to Run Agentic AI Across Its Entire Operation
Merck’s new up-to-$1-billion partnership with Google Cloud will embed agentic AI across drug research, manufacturing and corporate operations, joining a wave of pharma giants making nine-figure AI infrastructure bets in 2026.
Merck and Google Cloud announced on April 22, 2026, at Google Cloud Next in Las Vegas, a multi-year partnership worth up to $1 billion that will deploy Google’s Gemini Enterprise agentic AI platform across Merck’s research and development, manufacturing, commercial and corporate functions — one of the largest single AI infrastructure commitments any pharmaceutical company has made to date.
What “Agentic AI” Means Inside a Drug Company
Unlike a single chatbot or search tool, agentic AI systems are designed to independently carry out multi-step tasks — pulling data from several internal systems, running an analysis, drafting a document and routing it for human sign-off — with limited step-by-step human prompting along the way. Under the deal, Google Cloud engineers will work directly alongside Merck teams to build and deploy these agents across functions as varied as clinical trial operations, manufacturing quality checks, regulatory submission drafting and internal corporate workflows like finance and HR.
Why Merck Chose Google Over Rivals
The agreement puts Merck alongside a growing list of pharmaceutical giants placing large bets on a single cloud and AI partner rather than building everything in-house: Novo Nordisk struck its own strategic partnership with OpenAI and AWS earlier in 2026, while Eli Lilly has invested heavily in its own supercomputing infrastructure for drug discovery. Merck’s choice of Google Cloud centers on Gemini Enterprise’s ability to plug into existing Google Cloud data infrastructure Merck already uses, letting the company scale agentic workflows without a wholesale migration of its underlying data systems.
The Scale of the Commitment
The deal’s value — up to $1 billion over its multi-year term — reflects not just software licensing but the embedded engineering support model both companies emphasized in their announcement: Google Cloud staff working inside Merck’s own technology organization rather than simply selling a platform and walking away. That embedded-team approach has become increasingly common in large enterprise AI deals as companies discover that off-the-shelf AI tools often stall without dedicated integration engineering to connect them to messy, decades-old internal systems.
Where the AI Will Actually Show Up First
Merck executives have pointed to regulatory-document drafting and clinical trial data management as among the first areas expected to see agentic AI deployed at scale, given the enormous volume of structured and unstructured documentation those functions generate. Manufacturing quality and supply chain functions are also early targets, where agentic systems can flag anomalies across production data faster than manual review cycles that today can take days.
The Optimistic Case and the Skeptical One
Supporters of large pharma-AI infrastructure deals argue they are the only realistic way to compress the notoriously slow, expensive process of bringing a drug from discovery to approval, particularly as patent cliffs squeeze revenue across the industry and companies look for efficiency gains everywhere they can find them. Skeptics, including some inside the industry, caution that agentic AI systems operating across regulated functions like clinical trials and manufacturing quality carry real risk if they act on flawed data without adequate human oversight, and that a $1 billion price tag guarantees ambition but not necessarily results — pharma has a long history of expensive digital transformation projects that underdelivered relative to their announced scope.
What Happens From Here
Neither company has published a detailed public timeline for when specific agentic workflows will go live, but both have signaled the partnership is meant to run for multiple years rather than a single deployment cycle. With Merck now joining Novo Nordisk, Eli Lilly and other major drugmakers in committing nine- and ten-figure sums to AI infrastructure in 2026 alone, the pharmaceutical industry’s next competitive axis may be less about which company has the best drug pipeline and more about which one can most reliably turn its AI infrastructure spending into faster, cheaper drug development.
Photo: panumas nikhomkhai / PEXELS via Pexels
