1. Home
  2. Insights
  3. Blog
  4. Real-time clinical trials: a pivotal shift and a global question

Real-time clinical trials: a pivotal shift and a global question

Regulators have long sought to balance speed with scientific rigour in clinical development. Through its recent “Real-Time Clinical Trials” (RTCT) initiative and an associated pilot exploring AI-enabled optimisation in early-phase studies, the US Food and Drug Administration (FDA) has signalled a step-change in how it wants clinical development evidence to be generated and shared—shifting from periodic, sponsor-mediated reporting towards real-time visibility of key signals and more data-driven early decision-making early in development.1

However, while the FDA’s direction is gaining attention, a more important question for sponsors is emerging: how will this approach translate in a global regulatory landscape, where development and approval must extend well beyond the US to achieve commercial success?

Moving beyond the “lag problem”

At the centre of the FDA’s initiative is a well-recognised inefficiency: clinical trial data typically moves in stages, from sites to sponsors and only then to regulators, introducing delays that can stretch into months or even years. RTCT aims to reduce this lag by enabling near real-time sharing of trial data and key signals directly with the FDA.2 In parallel, the Administration is exploring how AI can improve early-phase decision-making, supporting areas such as dose optimisation, safety signal detection and recruitment efficiency.3

Importantly, this is not merely theoretical. The FDA has already initiated proof-of-concept trials, including oncology studies sponsored by AstraZeneca and Amgen, in which defined signals are transmitted and assessed in real time.1 Taken together, these efforts point to a broader vision: reducing the stop-start nature of traditional development and potentially enabling more continuous progression across phases but also raising important questions about how such a model could operate in a global regulatory environment.

From concept to reality: what changes for sponsors?

If scaled, RTCT could redefine “trial readiness” across three critical areas:

  1. Operational readiness: Sponsors will need robust, near real-time data pipelines and clear definitions of which signals trigger regulatory visibility
  2. Governance readiness: Greater transparency will place increased emphasis on control frameworks, including how AI models are validated, monitored and governed
  3. Evidence readiness: Will be essential to demonstrate that AI-driven or real-time approaches improve not just speed, but also the quality of decision-making2

However, feasibility alone is not the challenge, control is. Real-time visibility introduces new complexities in how data are curated, interpreted and shared, particularly when datasets may be less mature than those traditionally submitted for regulatory review.

It also shifts the dynamic between sponsor and regulator. Earlier, parallel visibility may reduce the sponsor’s opportunity to shape the narrative around a product, requiring greater confidence in protocol design and signal definition from the outset.

A global lens: convergence, divergence and “wait and see”

While the FDA’s direction is clear, global alignment is far less certain. Across major regulatory authorities including the European Medicines Agency (EMA), the UK’s MHRA, Japan’s PMDA and China’s NMPA there is strong momentum towards more data-driven and technology-enabled regulation. However, this momentum is largely directed towards adjacent capabilities rather than direct equivalents of RTCT.

In Europe, the EMA is advancing a structured framework for the use of AI across the medicine lifecycle, supported by a wider data and AI strategy running through to 2028. Regulatory expectations centre on the risk-based, transparent and validated use of AI, with sponsors expected to demonstrate robustness, explainability and data quality when deploying AI in clinical trials.4 Similarly, the UK’s MHRA is deploying AI and digital tools to improve efficiency and accelerate trial approvals, including systems that offer near real-time visibility of applications and support AI-assisted review processes.

However, these initiatives remain focused on enhancing regulatory decision-making, rather than enabling continuous, sponsor-driven data flow during trial conduct.

In Japan, the PMDA is expanding the use of real-world data and encouraging participation in multi-regional clinical trials, with strict requirements on data integrity, traceability and governance. This reflects a broader commitment to structured, globally harmonised evidence generation, rather than real-time regulatory interaction.5

China’s NMPA may represent the closest conceptual parallel. Its recent “AI + drug regulation” policy framework6 points to a long-term shift towards a more connected, data-driven regulatory ecosystem, incorporating continuous monitoring and lifecycle oversight enabled by AI. Although not specific to real-time clinical trials, this direction signals a move towards more integrated and continuous regulatory engagement.

Collectively, these developments point to convergence in strategic intent, faster, more data-driven decision-making, but continued divergence in implementation, with the FDA remaining the most explicit in piloting real-time regulatory visibility into ongoing trials.

The human factor: AI, oversight and trust

A consistent theme across regulators is the need for meaningful human oversight. AI is seen as an enabler of better decision-making, but not a replacement for expert human regulatory judgement or accountability.

Both EMA and MHRA frameworks emphasise transparency, accountability and explainability in AI use, alongside robust governance, validation and ongoing performance monitoring requirements. This is particularly relevant in a real-time environment, where data may be incomplete, evolving or subject to uncertainty and bias. Ensuring that signals are interpreted appropriately and that decisions are grounded in validated evidence remains essential.

In practice, this means AI-generated insights should support, rather than determine, regulatory conclusions with clear decision ownership, validation and escalation to protect trust, patient safety and regulatory integrity.

Designing for a divergent future

For global development programmes, the emergence of RTCT raises a strategic question: should trials be designed to meet the most progressive regulatory expectation, or adapted regionally?

The likely answer lies somewhere in between.

  1. Early-phase studies may increasingly incorporate real-time capabilities to support earlier engagement with the FDA
  2. Later-phase or global programmes may need to accommodate differing expectations on data maturity, submission readiness and review
  3. Protocol design will become even more critical, defining how and when data is shared, interpreted and governed across jurisdictions

Notably, global regulatory alignment efforts such as ICH guidance on adaptive trial designs recognise the value of more flexible, data-driven approaches, while still emphasising the need for rigour, transparency and prospective planning.7

What differentiates the FDA approach is the emphasis on continuous regulatory visibility into emerging trial data. This aligns closely with its long-standing “data-up” review philosophy, where regulators interrogate underlying datasets to derive conclusions. Some agencies, particularly in Europe, have historically adopted a more “top-down” approach, assessing whether submitted evidence supports a sponsor’s proposed label or claims.

This distinction matters. A real-time data feed may integrate more naturally into the FDA’s analytical model than into systems where evaluation pathways are structured differently. For these regulators, the shift to continuous data visibility could represent a more fundamental transformation, requiring changes not only to technology, but also to review processes, governance and decision-making rules.

The true challenge is therefore not simply technological adoption but designing systems that can operate effectively across different regulatory philosophies, evidentiary expectations and levels of readiness for real-time data review.

Looking ahead: from pilots to practice

The FDA’s RTCT initiative represents a meaningful shift towards more dynamic, data-driven regulatory engagement. Its success will depend not only on technical feasibility, but also on how well it balances speed with scientific certainty, regulatory control with sponsor-agency communications and early insight into evidentiary robustness.

Globally, the response is likely to be measured. While regulators are investing heavily in AI, digitalisation and real-world data, fully real-time regulator-facing trial models have yet to emerge at scale outside the US pilot environment.

For sponsors, the implication is clear: the future of clinical development is unlikely to be defined by a single regulatory model. Instead, success will depend on the ability to navigate and design for a more complex, more connected and increasingly data-driven landscape, where trial design, data governance and engagement strategy are considered from the start.

ICON continues to monitor these developments closely, helping sponsors translate emerging regulatory expectations into practical, globally viable clinical strategies.

 

References

  1. US Food and Drug Administration. FDA announces major steps to implement real-time clinical trials. April 29, 2026. https://content.govdelivery.com/accounts/USFDA/bulletins/414ec5f 
  2. Applied Clinical Trials. FDA Launches Proof-of-Concept Real-Time Clinical Trials. April 29, 2026. https://www.appliedclinicaltrialsonline.com/view/fda-real-time-clinical-trials 
  3. Federal Register. AI-enabled optimization of early-phase clinical trials pilot program request for information. April 29, 2026. https://www.federalregister.gov/documents/2026/04/29/2026-08281/ai-enabled-optimization-of-early-phase-clinical-trials-pilot-program-request-for-information 
  4. European Medicines Agency. Data and AI in medicines regulation to 2028 strategy. 2025. https://www.ema.europa.eu/en/documents/other/network-data-steering-group-workplan-2025-2028_en.pdf 
  5. Asano, J et al. PMDA Perspective on Use of Real‐World Data and Real‐World Evidence as an External Control: Recent Examples and Considerations. April, 2025. https://www.researchgate.net/publication/387709486_PMDA_Perspective_on_Use_of_Real-World_Data_and_Real-World_Evidence_as_an_External_Control_Recent_Examples_and_Considerations 
  6. NMPA. Implementation Opinions on “Artificial Intelligence + Drug Regulation”. April 2, 2026. https://www.nmpa.gov.cn/xxgk/fgwj/gzwj/gzwjzh/20260402091552114.html 
  7. ICH. Adaptive designs for clinical trials E20. June 25, 2025. https://database.ich.org/sites/default/files/ICH_E20EWG_Step3_DraftGuideline_2025_0625_0.docx 
In this section

Connect with us