Amgen CEO Reports 50% Faster Drug Candidate Selection Through AI Implementation
核心洞察
Amgen has achieved a 50% acceleration in drug candidate selection for clinical development through its comprehensive AI implementation strategy.
The company reduced manufacturing production line clearance time from 30 minutes to 2 minutes per batch and improved clinical trial enrollment efficiency by up to three times.
CEO Bob Bradway emphasizes that 2026 will be critical for implementing agentic AI systems to automate routine scientific tasks and free researchers for higher-value work.
Amgen has achieved a 50% acceleration in selecting drug candidates for clinical development through its comprehensive artificial intelligence implementation, according to CEO Bob Bradway. The biotechnology giant's early investment in AI technologies is now delivering tangible results across drug discovery, manufacturing, and clinical operations.
Manufacturing and Clinical Efficiency Gains
The company has demonstrated significant operational improvements through AI deployment. At one manufacturing site, AI has reduced production line clearance time from approximately 30 minutes to about two minutes per batch run. In clinical trials, AI tools have enabled participant enrollment up to three times more efficiently compared to traditional methods.
Amgen has implemented AI tools company-wide, including Microsoft Copilot (搜索) and OpenAI (搜索)'s ChatGPT Enterprise (搜索), alongside comprehensive training programs and responsible-use frameworks. These technologies are designed to be "additive and supportive" to research teams, focusing on freeing up time for scientists to concentrate on critical drug development activities.
Strategic AI Investment Timeline
Bradway's AI strategy began in 2012 with Amgen's acquisition of DeCODE Genetics (搜索), an Icelandic company holding longitudinal genomic data on virtually the entire population of Iceland. This acquisition represented a strategic bet that data and AI would eventually transform medicine discovery, despite initial skepticism from industry peers.
The company has since developed proprietary capabilities including protein (搜索) folding models and zero-shot antibody design, which uses AI to engineer drug molecules without prior experimental examples. An Nvidia (搜索) SuperPod now operates in Reykjavik, supporting these advanced computational efforts.
Rather than relying solely on external AI models like DeepMind (搜索)'s AlphaFold (搜索), Bradway pushed his team to identify specific capabilities and limitations of existing models, then build customized solutions. "We're not relying on any single approach," he stated.
Future Focus on Agentic AI
Bradway has identified 2026 as a critical year for implementing agentic AI systems. These advanced AI agents could automate routine scientific tasks such as form completion, material requisitioning, and data summarization, allowing researchers to focus on specialized scientific work.
"Scientists today burn hours filling out forms, requisitioning materials, summarizing data — work that agentic systems could absorb entirely, freeing researchers to do what only they can do," Bradway explained. He described this potential as "magic" and emphasized that AI should be "liberating, rather than displacing" for scientific staff.
Addressing Industry Skepticism
While acknowledging that "America broadly doesn't trust AI" and that uncertainty breeds anxiety among employees, Bradway expressed puzzlement at industry concerns about AI return on investment. He noted that in biotechnology, where single molecules can require decade-and-a-half development timelines, patience with new technologies is essential.
The CEO emphasized the importance of transparency in AI implementation to address employee concerns. "The more we can share transparently what we're doing, I think the better," he said, while acknowledging the challenge of managing uncertainty during technological transitions.
Competitive Implications
Bradway warned that delayed AI adoption could prove costly for biotechnology companies. "Being late to this party is going to be expensive," he stated. "Being early to the party has the potential for real benefit."
The traditional drug discovery process has historically relied on iterative laboratory work, manual data review, and sequential testing. AI enables researchers to analyze larger datasets, model molecular behavior, predict protein (搜索) structures, and evaluate candidate molecules earlier in development, with the goal of deprioritizing less promising candidates sooner in the process.