"The biggest gap and risk," she says, is the assumption that key-coded, pseudonymised patient data is not personal data at all. People think it’s therefore safe to use in AI tools without caution. That’s in fact a misconception.

In an industry quickly adopting generative AI for tasks from protocol drafting to drug discovery, this misunderstanding is becoming one of the costliest compliance oversights in clinical research.

In the recent episode of The Discovery Loop podcast, host Shubhangi Dua, Podcast Producer and B2B Journalist at Pharmatica.io, sat down with Diana Andrade, Founder & Managing Director at RD Privacy and Data Protection Officer (DPO) for Biopharma and Life Sciences. They talk about the current state of data protection regulations and how enterprises can innovate responsibly.

As the founder and lead at RD Privacy, Andrade serves as a data protection officer for biopharma and life sciences companies dealing with GDPR compliance, cross-border data transfers, and privacy governance across several jurisdictions.

Her career began in contracts at the contract research organisation (CRO) giant PPD clinical research business of Thermo Fisher Scientific. Eventually, she moved into privacy law, revealing a gap where many lawyers understood data protection, but few knew enough about clinical research to apply it correctly.

Also Read: 11 Clinical Trials Defining Pharmaceutical Strategy in 2026

What’s the AI Misconception Costing Pharma Companies

On the Pharmatica podcast with Dua, Andrade spotlighted two key AI misunderstandings she frequently sees among sponsors and CROs.

The first is technical. Teams confuse pseudonymization with anonymisation. Data that has been key-coded, stripped of clear identifiers but still traceable to an individual, is still considered personal data under GDPR. Using it in an AI tool does not change this classification, regardless of how the data is labelled internally.

The second is cultural and arguably more dangerous. Compliance teams are beginning to think AI can handle compliance on its own. "It's very hard with AI," Andrade says. AI adoption will continue to rise, and familiarity with these tools is now essential in the industry. However, this increases the need for strong governance.

Pharma enterprises still require major human oversight to establish guidelines and oversee AI's actions.

In practice, this requires internal processes defining how AI can be used and active human oversight of every output. Andrade told Dua that you can’t trust everything AI produces without verifying it first.

Where AI Has Earned Its Place in Clinical Research

Andrade acknowledges significant value in AI tech across the clinical trial lifecycle. For instance, AI’s benefits include faster drafting of clinical research protocols, developing internal procedures that help smaller biopharma teams scale without increasing headcount, and analysing trial data for new drug development leads.

However, the RD Privacy Managing Director is wary of using AI without supervision. She noted that each of these use cases involves feeding an AI system with sensitive material from business-confidential information and personal data, despite the data being key-coded. This combination is why she argues that pharma compliance programs must go beyond GDPR and specifically address emerging AI regulations, rather than treating these as separate issues.

Why Every Pharma Company Needs a Dedicated DPO?

Andrade advises pharma leaders that every biopharma enterprise needs a data protection officer or someone specifically responsible for its global privacy program, no matter its size.

The reason is the structure of clinical trials, which can’t be contained in one country. They start in one controlled area and then expand rapidly. The legal basis for processing patient data can entirely change when moving from one country to another. In some markets, patient consent is necessary; in others, the sponsor's legal obligations and legitimate interests come into play, needing a documented legitimate interest assessment to ensure these interests do not violate patients' rights.

Andrade believes that while compliance may come across as a puzzle, it can be managed with one piece but becomes exponentially harder as trials move into new markets. Each new country adds pieces that must fit together. Without a DPO to manage this puzzle, she warns, "very quickly your clinical trial can become a mess."

She emphasises that the DPO role should be viewed as an ongoing necessity, not a one-time task. It's not about filing documentation away and forgetting it; it requires active involvement that evolves with clinical operations and needs support from every growing team within the organisation.

Also Watch: AI in Pharma: Hype vs. What Actually Works

How to Stay Up-to-Date in Pharma Regulatory Space

To assist privacy professionals in keeping up with a regulatory environment that spans many jurisdictions, Andrade recently launched RD Privacy Watch. This platform aggregates and summarises data protection news and guidance from authorities worldwide, links to sources, and provides a weekly digest of the most relevant updates, which can be filtered by jurisdiction.

Andrade created it to address her own need to stay informed in a fragmented global regulatory market. Recognising its value to the broader privacy community, she made it publicly available.

She also runs the RD Privacy Global Academy, a training platform to help teams implement data protection controls in their daily work. She noted the training currently available is GDPR-focused, with plans to expand into other jurisdictions.

Takeaways

• AI is transforming data protection in clinical trials.
• Pseudonymized data still requires careful handling.
• Human oversight is essential when using AI.
• Global clinical trials face complex regulatory challenges.
• Training is crucial for compliance and empowerment.
• Data protection is an ongoing journey, not a one-time task.
• Privacy should enable innovation, not hinder it.
• Understanding local laws is vital for cross-border transfers.
• Trust in data protection is key for patient recruitment.
• RD Privacy Watch is a valuable resource for privacy professionals.

Chapters

• 00:00 Introduction to Data Protection in Pharma
• 04:00 The Impact of AI on Data Protection
• 11:50 Challenges in Cross-Border Data Transfers
• 17:28 The Importance of Training and Compliance
• 23:31 Launching RD Privacy Watch
• 30:16 Navigating Clinical Trials and Data Protection

For more information on data regulations for clinical trials globally, head over to rdprivacy.com and reach out to Diana Andrade on LinkedIn.

#ClinicalResearch #DataPrivacy #PharmaCompliance #GDPR #AI #Biopharma #ClinicalTrials #HealthTech #LifeSciences #DPO #DataProtection #TheDiscoveryLoop #RDPrivacy #RegulatoryAffairs #PatientData #AIinHealthcare #ComplianceStrategy

In this episode of Discovery Loop, host Trisha Pillay sits down with Christian Nkanga, Chief Scientific Officer at Memsel, to examine why the gap between scientific discovery and clinical reality remains so wide. They also covered what the pharmaceutical industry must do differently to close it.

Science Is Moving Faster Than the System Can Handle

Nkanga's perspective on this problem is shaped by more than professional expertise. As both a professor and a practising chief scientist, he has worked directly with diseases for which no adequate treatment exists. That experience gives his analysis a clarity that purely academic commentary rarely achieves. His starting point is a statistic that deserves more attention than it typically receives. Between 2000 and 2019, approximately two million publications were produced on nanotechnology. Of those, roughly 80 products received FDA approval. Two million studies, but only 80 approved products.

The science, Nkanga argues, is not the problem. The problem is the system through which that science must travel to reach patients. Regulatory requirements, safety validation, reproducibility standards, manufacturing scale-up, and quality management all stand between a laboratory result and a licensed therapy. Each stage creates another point where promising research can stall. The challenge for pharmaceutical professionals is not to dismiss these requirements. They exist for good reason. The challenge is to understand them early enough to design research programmes that can actually survive contact with them.

Documentation, Reproducibility, and the Trust Regulatory Agencies Require

Nkanga is equally direct about the internal practices that separate successful development programmes from those that fail to progress. Effective documentation and reproducibility are not bureaucratic obligations. They are the foundation on which regulatory trust is built. Regulatory agencies cannot approve what they cannot verify. When a research team submits data to support a clinical application, the quality of that documentation, its completeness, its consistency, and its internal coherence communicate something essential about the rigour of the science behind it.

Teams that treat documentation as an afterthought, to be assembled retrospectively once the scientific work is complete, consistently find themselves unable to satisfy the evidentiary standards that approval requires. Reproducibility carries the same weight. If a result cannot be replicated under defined conditions, it cannot form the basis of a product that will be manufactured at scale and administered to patients. Building reproducibility into research design from the outset, rather than hoping it emerges naturally, is one of the clearest predictors of whether a programme will survive into later development stages.

Translational Development as a Discipline

The broader argument running through this conversation is that translational development, the process of moving science from experimental success to clinical application, needs to be treated as a distinct and rigorous discipline in its own right. It is not a natural extension of good research. It requires a different set of skills, a different set of questions, and a different understanding of what success looks like.

For pharmaceutical professionals, that means building translational thinking into programmes from day one. It means asking not only whether a compound works in a controlled setting, but whether the evidence generated is sufficient to persuade a regulator, manufacturable at a viable cost, and relevant to the clinical problem it claims to solve. The patients waiting for those innovations do not have the luxury of waiting for the industry to work this out incrementally. The science is ready. The question is whether the systems, practices, and disciplines needed to translate it are ready too. For more pharmaceutical science and drug discovery insights, visit Memsel or connect with the guest:

Christian Nkanga: LinkedIn | Chief Scientific Officer, Memsel

Takeaways

• Innovation in drug discovery is accelerating, but real-world impact is limited.
• Only a small fraction of scientific knowledge translates into marketable products.
• Pre-research phases are critical and often underestimated in drug development.
• Successful programs focus on scalability and reproducibility from the start.
• Outsourcing non-core functions can help small companies focus on innovation.
• Good documentation practices are essential for regulatory approval and reproducibility.
• Engaging end-users early in the process is vital for successful product development.
• Milestone-driven contracts can mitigate risks in outsourcing.
• Building local expertise and mentorship is crucial for sustainable drug development.

Chapters

00:00 Introduction to Drug Discovery Challenges

03:02 Christian Nkanga's Journey and Insights

05:53 The Gap Between Innovation and Real-World Impact

08:59 Pre-Research Phase: Importance and Strategies

11:59 Factors Separating Successful Discovery Programs

15:45 Bridging Local Expertise Gaps in Drug Development

20:00 Key Lessons from Product Development Experience

27:03 Recommendations for Accelerating Drug Discovery

34:04 Conclusion and Call to Action