Drooid Logo
Back to story perspectives

Full Breakdown

AI-Driven Overhaul of GSK’s Drug-Safety Workflow

By Drooid · · How we work

Background and Rationale

GSK’s pharmacovigilance process—receiving safety reports via email, fax, paper and data loggers, then manually assessing priority and routing—can take hours and often requires repeated data entry. Lisa-Ann Chiriseri, director of drug safety and pharmacovigilance applications at GSK, describes the issue as “a data problem” rather than merely a process flaw. To address this, GSK has partnered with Cognizant, whose global head of intelligent enterprise platforms, Mohammad Haque, emphasizes that improving isolated tasks yields limited value when downstream teams must reconstruct or re-enter information.

Proposed AI Agent Model and Targets

The joint effort envisions a network of specialized AI agents that handle predictable, rule-based tasks (e.g., intake, coding, information checks) while human experts retain authority over interpretive decisions. Haque estimates that roughly 40 agents could eventually support the full drug-safety life cycle, enabling parallel work streams and reducing bottlenecks. Key performance targets include cutting overall case-processing time and the effort required for aggregate safety reports by 50 %, and shortening the interval from safety-signal identification to decision by 40 %. Cognizant also aims for greater than 90 % consistency in quality-control activities.

Official Statements & Responses

Chiriseri stresses that AI will not replace judgment: “We’re not asking for blind trust from our regulators and quality and compliance teams.” She notes that GSK is embedding audit trails to reconstruct how information is handled, and that validation and ongoing monitoring will keep systems within intended limits. Haque adds that success will be measured by outcomes—primarily time savings—and that early production use is already showing confidence-building results. Both executives highlight the importance of traceability, employee involvement through “citizen development,” and reusing existing data rather than rebuilding systems from scratch.

Verbatim Quotes

  • “We’re not asking for blind trust from our regulators and quality and compliance teams,” — Ann Chiriseri, director of drug safety and pharmacovigilance applications at GSK