人工智能增强的安全与药物警戒新视野：2026年专家预测

A New, AI-AugmentedHorizon for Safetyand Pharmacovigilance:Expert Predictions for 2026 IAN FISHER, Head of Development Analytics, Regulatory Affairs and Drug Developme nt SolutionsMICHELLE GYZEN, Senior Director, Strategic Regulatory Solutions; Head of Regulatory Services Innovation & TechnologyARCHANA HEGDE, Senior Director, PV Systems & Innovations, Lifecycle SafetyANA PEDRO JESUÍNO, Director of Marketed Product SafetyLOUISE MOLLOY, Director Medical Information & Pharmacovigilance Table of contents Introduction1Looking back at 2025, what was the most surprising development2in your area of expertise?What is the biggest misconception around AI’s role in compliance?3What are the critical capabilities regulatory intelligence tools should4integrate to support more harmonized compliance?The FDA’s draft guidance on AI in drug development has introduced4concepts like risk-based validation and context of use. What practical stepsshould PV teams take to operationalize these expectations in 2026?How should a team help prevent AI from making mistakes in safety reporting?5What confidentiality considerations are important when AI5is using and analyzing patient data?As AI and large language models become standard in PV and compliance,5what are some practical steps organizations can take to demonstratemodel validation, monitoring, and risk assessment to regulators?Are there any new oversight skillsets that local qualified persons for6pharmacovigilance (LQPPV) will need in 2026 as AI continues to becomemore embedded in safety workflows?What are the most important ways that teams should involve their6target product profiles (TPPs) to drive alignment, inspection readiness,and long-term value across development and commercialization?What transparency standards and safeguards should be in place7to ensure ethical use of AI in patient-facing interactions?How can smaller pharma companies keep up with all these changes?7What safety or regulatory trend will define 2026?8How should pharma teams prepare for it?About IQVIA8About the authors9 Introduction The question ofifAI can improve safety and regulatory workflows has beenanswered with a resoundingyes. Now, organizations must determine howto integrate this technology to maximize efficiency, ease the burden on theirteams, and ensure regulatory compliance. In a recent panel discussion,five IQVIA experts sat down to share their predictions for how safety andregulatory workflows will evolve in 2026, providing their expert guidance onthe importance of keeping a human-in-the-loop, the role of AI governance inmaintaining compliance, and how best to train staff for this major technologyshift. New trends are on the horizon: leverage these expert insights to helpyour organization ride the AI wave to strategic success. “Health authorities are asking notjust if you’re using AI but how you’regoverning it. This has forced a lot oforganizations and regulatory teamsto evolve their tech fluency in 2025.” Looking back at 2025, whatwas the most surprisingdevelopment in your areaof expertise? Archana Hegde, Senior Director, PV Systems & Innovations, Lifecycle Safety:The most surprisingaspect is the cultural change we are seeing amongst theprocess owners, CROs, and end users of technology forpharmacovigilance (PV). We’ve seen a shift in terms oftheir participation, from not being able to digest all thetechnology to providing input on developing technology.The second is regulatory responses from the FDA, EMA,MHRA (the UK’s Medicines and Healthcare productsRegulatory Agency), etc. They’ve given us guidance onhow to use AI and what our frameworks should look like. — Michelle Gyzen Louise Molloy, Director Medical Information & Pharmacovigilance:With advancements in technology,there’s been a repositioning of the value of medicalinformation. It used to be perceived as an add-onservice. Now, it’s being recognized as a source ofstrategic insights. In gathering medical information, wehave direct contact with patients, HCPs, and caregiversthat provide real-world insight. Over the past year,people have been exploring how to use that insight toinform labeling and ensure it is fit for purpose. The voiceof the consumer is being brought into the conversation. Ian Fisher, Head of Development Analytics,Regulatory Affairs and Drug Development Solutions: We’ve seen an increase in the transparency aroundhow regulatory decisions are made. One big surpriseis the FDA’s release of theComplete Response Letters(CRLs) and the clarity in terms of: these are the hurdlespeople have had to face to make it to market. If usingAI tools, especially agentic AI, and asking questionsfrom different perspectives, we can draw out valuablenuggets of information from CRLs, summary bases ofapprovals, and EPARs (European Public AssessmentReports) to increase success rates. Ana Pedro Jesuíno, Director of Marketed Product Safety:It was the rapid integration of AI-driven toolswithin PV scope and oversight to increase efficiencyand o