Determining RoI for Generative AI use cases to gain funding: Complexities and Solutions

From Hype to Action

Generative AI (Gen AI) is no longer just a futuristic concept in pharmacovigilance—it's a tool that's already reshaping how safety data is processed and managed. But for many organizations, the path to integration of AI in pharmacovigilance remains unclear. To explore how PV leaders are navigating this shift, the Indegene's Pharmacovigilance Digital Council sponsored this survey-based report, gathering insights from industry experts to uncover the priorities, challenges, and opportunities in adopting Gen AI solutions.

Based on the survey findings, this paper highlights emerging trends in in AI in decision-making, key performance indicators (KPIs), ROI measurement challenges, change management, communication strategies, and barriers to securing funding for AI in regulatory compliance. Among the notable insights:

• Efficiency gains emerged as the most critical KPI, signalling a strong focus on automation and streamlined workflows.
• Decision-making for Gen AI investments is increasingly a collaborative effort between IT and business teams.
• Justifying Gen AI ROI remains the biggest hurdle in securing funding, emphasizing the need for robust business cases.
• Planning for change management early in AI projects is crucial to address challenges and ensure smooth implementation. However, it is often treated as an afterthought, which can negatively impact project success.

By diving into these findings, this report provides actionable recommendations and strategies to help Pharmacovigilance organizations harness the full potential of Gen AI impact on pharmacovigilance and regulatory functions.

Efficiency, Savings, or Accuracy: What Matters Most?

The percentages represent the proportion of respondents choosing each option. Total Respondents: 10

Efficiency gains stood out as the most critical KPI, with 60% of respondents prioritizing it as a measure of success for Gen AI impact in pharmacovigilance initiatives. Close behind, cost savings were selected by 50% of respondents. Efficiency gains reflect the promise of Gen AI in pharmacovigilance to automate and streamline labour-intensive processes such as adverse event (AE) intake, case processing, and regulatory reporting. By reducing manual intervention, PV teams can save time, accelerate timelines, and reduce the risk of bottlenecks. In a domain where timeliness directly impacts patient safety, compliance is a given and expected as a default. With regulatory adherence already embedded into operations, efficiency becomes the defining metric for success, ensuring that automation drives both speed and precision without compromising compliance.

Cost savings, on the other hand, resonate with leadership teams striving to do more with constrained resources. The cost of manual PV operations—especially in scaling teams to handle growing AE volumes—is unsustainable. Gen AI ROI offers a compelling alternative: automating repetitive tasks to free up valuable human resources for strategic decision-making, thereby delivering tangible ROI.

Image-1: 5 Key Strategies for Driving Efficiency and Cost Savings with Gen AI

Who's Driving the Gen AI Revolution in PV?

The percentages represent the proportion of respondents choosing each option. Total Respondents: 10

When asked who drives decision-making for Gen AI investment and impact in pharmacovigilance, 37% of respondents identified collaboration between IT and business teams as the leading approach. This closely aligns with the 36% of respondents who indicated IT departments as the primary decision-makers.

The near-equal split between collaborative and IT-led decision-making underscores a critical shift in the industry. Historically, IT teams have driven technology adoption, leveraging their expertise in systems and infrastructure. However, as Gen AI in pharmacovigilance directly impacts core PV operations—such as adverse event processing and safety reporting—business and operational teams have stepped up to ensure alignment with process outcomes and compliance requirements.

The new age mandate for PV leaders is

• Break Down Silos for Better Outcomes — to ensure solutions are both technically robust and operationally relevant.
• Balance Technical and Operational Needs — to maximize Gen AI impact.
• Empower Cross-Functional Leadership — to guide the adoption process, ensuring alignment across departments.

ROI Isn't Just a Number—It's a Narrative

The percentages represent the proportion of respondents choosing each option. Total Respondents: 10

When asked about the primary challenges in calculating the ROI of Gen AI initiatives in pharmacovigilance, 30% of respondents highlighted difficulty in tracking accurate data as the most significant barrier. The difficulty in tracking accurate data can lead to unreliable insights and misaligned ROI calculations, leaving PV leaders uncertain about the true impact of their Gen AI investments. For PV teams, this means establishing robust data management systems should be a top priority. Clear guidelines on data capture, normalization, and integration across systems are essential to provide reliable benchmarks for ROI analysis. Without these systems in place, it becomes nearly impossible to measure improvements in case processing times, cost savings, or efficiency gains accurately.

Closely following data accuracy, 27% of respondents mentioned a lack of established frameworks as a critical issue. This suggests that many organizations are still in the early stages of defining clear metrics and methodologies for assessing Gen AI's impact. Without a structured framework, organizations risk evaluating Gen AI success on inconsistent or subjective criteria, making it difficult to compare results across teams or projects.

Here is a Toolkit for Building a Gen AI ROI Framework in PV

What's Holding Back Gen AI's Full Potential?

The percentages represent the proportion of respondents choosing each option. Total Respondents: 10

The most significant hurdle, identified by 36% of respondents, is justifying Gen AI ROI to stakeholders, followed by data privacy concerns (18%), which ties with other notable challenges like high setup costs and proving scalability.

Securing funding often hinges on a clear demonstration of value. The high prevalence of ROI justification as a barrier reflects the complexity of quantifying Gen AI's benefits, which often span multiple operational areas, including efficiency, quality, and compliance. Many PV leaders struggle to translate these gains into tangible metrics that resonate with decision-makers. For instance, while Gen AI may reduce case processing time, this must be framed in terms of cost savings, compliance risk mitigation, or enhanced patient safety to make a compelling case.

Adopting a business-driven approach could be the way forward:

1. Build a compelling business case by linking operational metrics to financial outcomes. For example, if Gen AI reduces manual labor by 20%, translate this into dollar savings on resources.
2. Use pilot projects to demonstrate measurable impact, such as time saved in literature review or improved accuracy in adverse event detection, to build credibility.

Data privacy concerns, cited as a significant barrier to Gen AI adoption, are not merely about compliance—they are about building trust across internal teams, stakeholders, and regulators. The sensitive nature of pharmacovigilance (PV) data demands a proactive, layered approach to security, especially when dealing with large-scale data aggregation or external vendors. Stakeholders are hesitant to invest unless convinced that privacy risks are mitigated effectively and transparently.

What Practitioners can do about data privacy in Gen AI?

1. Secure-by-Design Infrastructure

• Use Gen AI platforms with privacy as a core feature: end-to-end encryption, multi-factor authentication, and audit trails.
• Conduct regular security assessments and penetration tests to identify vulnerabilities, ensuring compliance with healthcare regulations such as GXP, 21CFR, GAMP 5, HIPAA or GDPR.
• Restrict sensitive data access through role-based access controls (RBAC) to ensure regulatory compliance.

2. Privacy-Integrated Workflows

• Anonymize patient data early and limit the use of personally identifiable information (PII) in training datasets.
• Establish robust data lifecycle management with secure storage, timely deletion, and encrypted backups.

3. Transparency and Stakeholder Trust

• Share updates on compliance measures, audit results, and certifications like ISO 27001.
• Train teams on privacy best practices and regulatory requirements to foster accountability.

4. Strategic Vendor Collaboration

• Rigorously evaluate vendors' security measures and compliance capabilities.
• Define clear responsibilities and breach protocols through data-sharing agreements.

5. Balancing Privacy with Functionality

• Work with IT to design scalable solutions that protect data without limiting Gen AI's usability.

How Big Are the Savings? PV Leaders Weigh In

The percentages represent the proportion of respondents choosing each option. Total Respondents: 10

The survey findings reveal a cautiously optimistic perspective on the cost-saving potential of Gen AI in pharmacovigilance. A significant portion of respondents (27%) believe that Gen AI impact can realistically achieve cost savings in the range of 10-20%, while an equal proportion anticipate savings in the more ambitious 21-30% range within the next 1-3 years.

These insights reflect two key trends:

• Realistic Optimism in Early Adoption: The 10-20% cost savings projection suggests that many PV leaders view Gen AI in pharmacovigilance as a valuable tool for incremental improvements rather than a disruptive force. This aligns with its current applications in tasks such as automating case triage, literature monitoring, and adverse event detection. These processes, while critical, represent areas where efficiency gains are relatively straightforward to quantify, making them ideal starting points for Gen AI implementation.
• Confidence in Strategic Gains: The optimism around the 21-30% savings highlights a growing belief in Gen AI's scalability and strategic value. As organizations refine their use of AI-powered tools, the focus shifts from immediate gains to broader efficiencies across the pharmacovigilance lifecycle. For example, using Gen AI to improve data harmonization, enhance signal detection accuracy, or streamline regulatory submissions could unlock this level of savings.

Conclusion: Your Roadmap to Gen AI Success

Generative AI is no longer a distant possibility—it's a transformative force reshaping pharmacovigilance today. However, realizing this potential requires more than just enthusiasm. It demands collaboration between IT and business teams, clear ROI narratives to secure stakeholder buy-in, and robust strategies to overcome hurdles like regulatory ambiguity and data privacy concerns. The question now is not whether Gen AI impact will change pharmacovigilance, but who will lead this transformation. By taking decisive action—piloting small, measurable initiatives and aligning technology with organizational goals—PV leaders have the opportunity to redefine operational excellence.

The path is clear, the tools are available, and the stakes are high. The time to lead is now.

Members of the Pharmacovigilance Digital Council

Fatima Bhayat MD, MPH, PhD, FFPM

Council Chair, Indegene

Dhaval Panchal

VP, Head Safety Surveillance, Risk Management and Clinical Safety, Viatris

Sarah Freestone

VP, Head of PV& Device Safety, UCB

Michael Murphy

Associate VP, Head of PV Operations, Amgen

Tjark Reblin

SVP, Head of Global Patient Safety, Ipsen

Cibele Vieira Cunha Rudge

Director of Single Case Processing, Haleon

Dr. Amrit Gill, MD MBA

Global Head of Consumer Safety – GMPV, Chattem, Inc.