AI in MedTech: Why Digital Foundations Will Determine Who Gets Ahead

AI in MedTech: Are Manufacturers Ready?

AI is creating significant expectations across the medical technology industry. From reducing administrative workloads to identifying potential safety signals earlier, its applications could change how medical device businesses operate. 

But there is a practical question manufacturers need to address before they can realise those benefits: are their existing systems and data ready to support it? 

For companies still relying on spreadsheets, paper records or disconnected software, introducing AI may prove more difficult than expected. The technology can only work effectively when it has access to reliable information, appropriate controls and a clear understanding of how the business operates. 

At the Veeva MedTech Summit in Amsterdam, Annemien Pullen, Vice President, MedTech Strategy, Europe, and Seth Goldenberg, President, Veeva MedTech, discussed why the industry’s structure, regulatory environment and digital maturity will influence how quickly manufacturers can make use of AI. Their observations point to a wider challenge: the companies best positioned to benefit may be those that have already invested in their digital foundations.

1. Why AI Presents a Different Challenge for MedTech

Medical technology has a distinct operating model compared with pharmaceuticals. Medical device manufacturers often manage large product portfolios, numerous configurations and frequent product changes, sometimes with relatively lean teams. 

As Goldenberg explained, a medical device business may have hundreds or thousands of product variations to manage, while continuing to develop and update existing products. This creates a substantial volume of information and activity across manufacturing, quality, regulatory and commercial functions. 

The challenge is not simply managing more products but keeping information accurate and connected as those products evolve. 

A change to a device may affect its components, manufacturing processes, technical documentation, regulatory status and information supplied to the market. When those activities are managed across separate systems, teams can spend considerable time finding information, checking records and coordinating updates. 

This is where AI could offer practical value. Rather than focusing on a single application or product type, its potential lies in helping teams manage operational complexity: reducing repetitive tasks, improving access to information and supporting decisions across business functions. 

One example discussed by Goldenberg was complaint handling. Better connected information could help manufacturers process complaints more efficiently and identify potential safety signals earlier, giving teams more opportunity to investigate issues before they develop into larger problems. 

However, these benefits depend on the quality and accessibility of the information behind them.

2. Connected Data is the Starting Point

AI tools need reliable, relevant information, supported by systems that understand the rules governing how that information can be used. 

Pullen highlighted the importance of establishing a sound data foundation and ensuring that core business applications provide the rules and controls AI needs to operate appropriately. 

Without this, an AI-generated response may be inaccurate, incomplete or unsuitable for a regulated environment. There are also questions around confidentiality and access: employees should only be able to access information they are authorised to see, and AI tools need to respect those same restrictions. 

For medical device manufacturers, this makes the wider digital infrastructure particularly important. If information is scattered across spreadsheets, paper records and disconnected applications, introducing AI on top of those systems will not automatically resolve the underlying problems. 

Product lifecycle management provides one example of why connected information matters. 

PLM acts as a central source of product information, including components, configurations and manufacturing details. When connected with other business functions, it can help manufacturers maintain visibility of changes throughout a product’s lifecycle. 

Consider a complaint received from the field. It may trigger a quality investigation, which identifies the need for a product change. That change could then require input from engineering and regulatory teams before updated information is communicated back to the field. 

Each stage depends on accurate information being available to the next team. If the process is fragmented, organisations risk delays, duplicated work and gaps in their understanding of the product’s current status. 

Goldenberg also highlighted the importance of maintaining what he described as change fluency. Historically, some manufacturers have managed product changes as they arise, gathering the relevant information when a regulatory submission becomes necessary. That approach can become increasingly difficult as regulatory expectations and product portfolios evolve. 

Maintaining an up-to-date picture of product changes makes it easier to understand their implications and respond when action is required.

3. Europe’s Digital Transformation Challenge

For European manufacturers, the case for digital investment is shaped by a demanding regulatory and commercial environment. 

Pullen pointed to the cost and resources involved in maintaining products on the European market, particularly as companies navigate requirements including the MDR, IVDR and the EU AI Act. 

These pressures make business cases for new technology particularly important. Manufacturers need to understand what an investment will deliver, whether through improved efficiency, reduced administrative work, faster processes or better access to information. 

There is also a difference in the starting point for some European organisations. 

Pullen observed that companies moving from paper-based processes and spreadsheets to cloud-based systems may face a different challenge from those replacing existing software. The change affects how people work, how information is managed and how teams interact with one another. 

That means digital transformation requires more than selecting and implementing technology. Organisations need to prepare their people, review existing processes and give teams the support to adopt new ways of working. 

The pace of technological development can make this difficult. New capabilities continue to emerge, but their value depends on whether organisations can incorporate them into everyday operations.

4. Can Smaller Manufacturers Keep Pace?

The growing use of AI also raises questions about the gap between large, digitally mature businesses and smaller medical device companies. 

Larger organisations may have dedicated technology teams and the resources to develop their own AI capabilities. Smaller manufacturers often have fewer people and less capacity to manage complex digital transformation projects. 

Goldenberg suggested that this could widen the difference between companies that have established strong digital foundations and those that have yet to begin that process. If AI helps some businesses save days or weeks of work, the cumulative advantage could become significant. 

For smaller manufacturers, the challenge is finding a proportionate starting point. Digital transformation does not have to mean implementing every available system or attempting to automate every process at once. 

A more practical approach is to identify where information gaps, repetitive tasks or disconnected workflows create the greatest operational difficulties. Addressing those areas can help establish a stronger foundation for future investment. 

The important consideration is whether the technology supports the company’s actual needs and can develop as the business grows. 

5. Why People and Processes Still Matter

Even with more capable AI tools, human judgement will remain essential in medical technology. 

Goldenberg discussed how AI could reduce repetitive work in areas such as medical, legal and regulatory review, allowing specialists to spend more time on complex decisions, new claims and higher-value activities. However, he emphasised that human oversight would remain important. 

Pullen also highlighted a challenge that can be overlooked amid the enthusiasm around AI: some organisations are simply not ready to adopt it. 

If the underlying data, applications and processes are not in place, introducing AI too early could result in significant investment without meaningful returns. In those circumstances, taking a step back and addressing the foundations may be the more productive decision. 

That requires organisations to be realistic about their current capabilities. Before investing in AI, manufacturers should consider three questions: 

• Is our information reliable and accessible? Can teams find the product, quality and regulatory information they need, and is it sufficiently accurate and up to date? 
• Are our systems connected? Can information move between functions without relying on manual intervention or repeated data entry? 
• Are our people and processes ready? Do employees understand how new tools will fit into their work, and are the necessary controls and responsibilities clear? 

AI has considerable potential to improve how medical device businesses operate. But the benefits will not arrive simply by introducing another tool. They will depend on the quality of the information, systems and processes supporting it. 

For manufacturers, building those foundations is a practical step towards making AI useful, manageable and sustainable as the technology continues to develop.