Many pharmaceutical companies still operate with fragmented content ecosystems, inconsistent governance structures, and reference materials that have evolved organically over many years.
Reference libraries often contain:
- Duplicate documents
- Outdated materials
- Inconsistent metadata
- Variable naming conventions
- Limited traceability across assets
Similarly, claims libraries may not always align with current campaigns, provide clear links to supporting evidence, or exist in formats that can easily be interpreted by AI systems.
These issues may be manageable in a heavily human-driven environment, where individuals use experience and judgement to navigate imperfections. AI systems are less forgiving.
If the underlying data is incomplete, inconsistent, or poorly governed, organizations risk automating inefficiency rather than eliminating it. The reality is that agentic MLR depends on trusted foundations. AI can only be as effective as the information that supports it.
AI readiness as a new capability
As organizations begin exploring agent-led review models, AI readiness is emerging as a new operational capability.
Successful implementation requires more than deploying technology. Organizations must establish processes, governance structures, and content standards that allow AI to operate effectively.
This includes:
- Maintaining high-quality reference libraries
- Ensuring claims are linked to supporting evidence
- Creating content in AI-readable formats
- Establishing governance for AI-generated outputs
- Defining ownership and accountability models
- Managing AI usage rights and permissions
These activities represent significant investment and ongoing operational effort. They require new ways of working and, in some cases, entirely new responsibilities across the organization.
The organizations that approach AI purely as a technology deployment often underestimate the foundational work required for long-term success.
The critical role of human judgement
While AI capabilities continue to evolve rapidly, there remains an important distinction between automation and judgement.
AI performs exceptionally well when applied to structured tasks, pattern recognition, and rule-based activities. It can process large volumes of information far more quickly than humans and identify issues that might otherwise be overlooked.
However, MLR decisions are not always governed by simple rules.
Many reviews involve interpretation, context, and professional judgement. Regulatory requirements differ between countries and regions. Scientific nuances may influence how information should be presented. Risk tolerances vary depending on the communication, audience, and market.
These factors often require experienced professionals to make informed decisions in situations where there may be no single definitive answer.
This is why the future of MLR is unlikely to be fully automated.
Instead, the most effective models will combine the efficiency of AI with the expertise, accountability, and judgement of experienced Medical, Legal, and Regulatory professionals. Regulatory bodies are also reaching a similar conclusion. In the UK, the PMCPA states that while AI may be used to support compliance activities, it does not absolve companies of their responsibilities under the ABPI Code, and requirements such as final certification by a nominated signatory cannot be fulfilled solely through the use of AI [pmcpa.org.uk].
In this model, AI provides scale and consistency, while humans retain responsibility for oversight, interpretation, and final decision-making. Rather than replacing MLR teams, AI has the potential to elevate them by reducing administrative burden and allowing experts to focus on higher-value activities.
Transforming more than technology
The adoption of agentic AI also requires broader organizational transformation. People need confidence that AI is operating reliably, transparently, and within clearly defined boundaries.
Technology alone will not address underlying process inefficiencies. Existing workflows often need to be redesigned to take full advantage of AI-enabled capabilities.
Standard operating procedures must evolve to reflect new decision points and responsibilities. Teams require training to understand how AI outputs should be interpreted and applied. Governance frameworks need to clearly define where automation is appropriate and where human review remains mandatory.
Perhaps most importantly, organizations must invest in change management.
A practical path forward
The future of MLR is not a choice between humans and AI. It is about creating an effective partnership between the two.
Organizations that succeed will take a phased and pragmatic approach. They will focus first on establishing strong data foundations, modernizing content governance, and creating AI-ready operating models. Pilot programs can then be used to test, learn, and refine approaches before broader deployment.
The organizations that derive the greatest value from agentic AI will not necessarily be those that adopt it first. They will be those that prepare most effectively for its use.
As content volumes continue to grow and expectations for speed and personalization increase, AI will undoubtedly reshape the future of content approvals. However, sustainable transformation will depend on more than sophisticated algorithms.
It will depend on the quality of the underlying data, the strength of governance frameworks, and the ability to combine technological efficiency with human expertise.
The future of MLR is not fully automated. It is thoughtfully augmented. It brings together the speed and scale of AI with the judgement, accountability, and scientific rigor that only experienced professionals can provide.
Want to know more about how IQVIA could help? Contact us below.
.svg){kind=icon}
