BMS deploys Anthropic's Claude across 30,000 employees to accelerate drug discovery and tackle data fragmentation, in one of pharma's largest enterprise AI rollouts.
Bristol Myers Squibb announced Wednesday it will deploy Anthropic's Claude to more than 30,000 employees, one of the largest enterprise AI rollouts in the pharmaceutical industry. The partnership targets the full drug pipeline: discovery, clinical development, and commercial operations.
The scope goes further than a standard enterprise license. BMS says it is collaborating directly with Anthropic to shape how Claude Code, Anthropic's coding assistant, gets integrated into research and business workflows. Greg Mayors, BMS's chief digital and technology officer, described the ambition in terms that signal the company is not interested in incremental gains: most enterprise AI stalls at the chat interface, he said, while the real value remains locked inside decades of siloed data.
That argument is essentially a bet on agentic artificial intelligence. Per a McKinsey estimate cited by The News, agentic systems requiring minimal human intervention could increase clinical development productivity by 35% to 45% over the next five years. Deploying Claude across 30,000 seats is BMS's structural move to be positioned when those gains materialize.
The competitive context
BMS is not alone. Eli Lilly has teamed with Nvidia to improve success rates in clinical trials, reflecting a broader wave of pharmaceutical AI partnerships. The race to prove that artificial intelligence in medicine can do more than summarize research literature is accelerating fast. Whether any of these deals eventually produce approved therapies is a separate question; most high-profile drug discovery AI collaborations announced over the past three years have not yet cleared that bar.
Anthropics has been more selective about enterprise announcements than OpenAI, which has moved aggressively to broaden ChatGPT into commercial workflows. Digital Watch Observatory noted earlier this year that AI providers face mounting pressure to demonstrate revenue beyond venture capital injections and speculative partnerships. The BMS deal hands Anthropic a concrete, regulated-industry reference account at meaningful scale.
The Claude Code dimension deserves specific attention. Most enterprise AI pilots treat developer tooling as a separate initiative from general employee productivity. BMS appears to be collapsing that distinction, exploring how coding assistance can connect computational drug modeling to downstream lab work. If it holds, that integration could compress the gap between in-silico predictions and bench validation, which matters in an industry where a single trial failure can cost hundreds of millions of dollars.
The data problem
Pharmaceutical R&D generates sprawling volumes of clinical, genomic, regulatory, and manufacturing data that have historically lived in incompatible systems. Mayors' reference to "decades of data silos" points directly at this structural barrier. Generative AI tools are only as useful as the data they can access, and BMS appears to be treating the Anthropic partnership as a forcing function to break down that fragmentation rather than just acquire a license.
The timing lands in a crowded week for AI news. Google, at its I/O 2026 event on Tuesday, announced the Gemini 3.5 model family and unveiled Gemini Spark, a cloud-based AI agent that works across apps in the background, per TechCrunch. Google now claims 900 million monthly Gemini users. The enterprise AI market is hardening into a real competition between well-resourced platforms, and large companies are picking sides.
McKinsey's productivity projections remain estimates, not evidence. Demonstrating that AI delivers measurable clinical efficiency at the scale BMS is targeting will take years, and the company is making a pipeline-wide commitment at a moment when the industry is still determining what AI can reliably do in a regulated environment. Mashable's coverage of I/O captured the broader pattern: every major platform is claiming the shift from passive assistant to active agent is now underway. BMS is stress-testing whether that claim holds up in one of the most demanding operational environments imaginable.
The next benchmark will be whether Anthropic and BMS can surface productivity data before the next wave of pharma AI partnerships resets expectations again.
---
Frequently Asked Questions
What will Bristol Myers Squibb use Claude AI for?
BMS plans to deploy Claude across drug discovery, clinical development, and commercial operations for more than 30,000 employees. The company is also working with Anthropic to integrate Claude Code into research workflows.
How does this deal compare to other pharma AI partnerships?
Eli Lilly partnered with Nvidia for drug success-rate modeling, but the BMS-Anthropic deal is notable for its breadth: it covers the full development pipeline and includes co-development of coding tools, not just access to a model.
What is Claude Code and why does it matter here?
Claude Code is Anthropic's AI-powered software development assistant. In a drug development context, it could help researchers automate computational modeling tasks and connect them more tightly to experimental workflows.
How much could AI improve pharmaceutical R&D productivity?
McKinsey estimates that agentic AI systems could improve clinical development productivity by 35% to 45% over the next five years, though that figure is a projection based on current trajectories, not a measured outcome.