Cencora高管谈制药业AI代理是否推进过快：AI应建议而非决策，需先定代理策略

Cencora’s Matt Sample and Gregg Gorniak explain why AI should recommend, not decide, and why pharma needs an agent strategy before chasing use cases.

Response to shifts in drug demand often remains reactive and siloed, according to Matt Sample, senior vice president of manufacturer, quality and replenishment operations at Cencora. A distributor seeing a high volume of omits might work with the manufacturer, or a customer facing an omit might contact the manufacturer directly. “It’s all done in our silos,” Sample said.

Speaking with Pharmaceutical Commerce at LogiPharma USA 2026, alongside Gregg Gorniak, vice president of manufacturer operations and data services and secure supply chain lead at Cencora, Sample described a goal of building a unified view of demand. It would pair end-to-end visibility into demand peaks and risks with data fed back to manufacturers. Sample said it is a concept Cencora wants to implement next year. The distributor could ask customers to validate forecasts for their top 10 SKUs, then share a six-month outlook with manufacturers to inform production planning. Sample noted that a historical trust factor shapes how much data customers, manufacturers and distributors share. “But at the end of the day, what’s right for the patients is having that open channel of data sharing and communication to drive the best patient access,” he said.

Sample said AI is still in its infancy when it comes to digitizing business processes and knowledge. The goal is to capture the decisions of the most experienced buyers so AI can recommend actions when an issue arises, with a human in the loop on everything. Handing AI full purchasing authority is a different story. “That’ll get messed up, not if, but when,” he said. He added that setting goals for AI agents, measuring their performance and course-correcting them will be a business role, not a technology role.

Gorniak agreed that AI needs human monitoring. He said some manufacturers he met with are already discussing their agents talking to other companies’ agents and making decisions. “They’re running before they can walk,” he said, adding that he wants AI to deliver data-based recommendations before it makes decisions. Sample cautioned that companies are pursuing AI use cases without first setting an agent strategy, which risks duplicated work. “The joke is we have 10,000 dashboards today all saying the same thing,” he said.

The pressure for earlier warning is not easing. ASHP and the University of Utah Drug Information Service counted 87 new drug shortages through the third quarter of 2026.¹ A Gartner survey of 509 supply chain leaders found that changes in ways of working driven by AI and agentic AI was the single most influential driver reshaping supply chain strategy over the next two years.²

References

1. American Society of Health-System Pharmacists; University of Utah Drug Information Service. National Drug Shortages: January 2001–September 2026. 2026. https://www.ashp.org/-/media/assets/drug-shortages/docs/Drug-Shortages-Report.pdf
2. Gartner Survey Shows 55% of Supply Chain Leaders Expect Agentic AI to Reduce Entry-Level Hiring Needs. Gartner. Feb. 25, 2026. https://www.gartner.com/en/newsroom/press-releases/2026-02-25-gartner-survey-shows-55-percent-of-supply-chain-leaders-expect-agentic-ai-to-reduce-entry-level-hiring-needs

Related to this article