Real-Time Clinical Trials: Maxis AI Meets FDA’s Shift.

The FDA is formalizing real-time clinical trials. The harder question is what clinical teams do after the signal arrives.

Real-time clinical trialsare not only about faster access to data. They create a new operating requirement: clinical teams must convert real-time signals into governed, traceable, human-supervised action. In April 2026, the FDA made that requirement explicit, taking concrete steps to let its reviewers see trial data as it is generated. It is a shift Maxis AI has been building toward for some time – the governed, real-time execution FDA is now encouraging is the operating model Maxis already runs for Sponsors and CROs.Across clinical development, visibility keeps improving: more data, more dashboards, more real-time feeds. What has not kept pace is the capacity to act on all of it, reliably and under control. Execution capacity, not visibility, is what decides who benefits from real-time trials.

Key Takeaways

The FDA’s April 2026 real-time clinical trials initiative shifts regulatory engagement from delayed, periodic review toward continuous, in-study visibility.
Real-time visibility creates a new operating requirement: teams must convert signals into governed, human-supervised, traceable action.
Systems of record (EDC, CTMS, eTMF, safety, lab) store and display data but do not coordinate execution across teams.
A governed AI Workforce supplies the missing execution layer, turning real-time signals into supervised, traceable action within existing systems.
Governance, not speed alone, determines what scales: role-based access, human validation, audit trails, and reproducibility.

In this article, you will learn

What Are Real-Time Clinical Trials?
Why the FDA’s RTCT Move Matters
The Real Challenge Is Not Visibility Alone
From Real-Time Data to Real-Time Execution
Where Sponsors and CROs Will Feel the Pressure
Why Existing Systems Are Necessary but Not Enough
The Role of a Governed AI Workforce
Governance Will Determine What Scales
The Future: Continuous Oversight, Continuous Execution
Frequently Asked Questions

What Does the FDA’s Real-Time Clinical Trials Initiative Actually Means?

Real-time clinical trials allow selected endpoints, safety signals, and other trial data to be reviewed as a study progresses, rather than only after traditional reporting and submission cycles. On April 28, 2026, the FDA announced two successful proof-of-concept real-time trials in which sponsors report endpoints and data signals to the agency as they occur, instead of collecting, analyzing, and submitting data in batches.¹ The agency described this as a step toward “continuous” trials, in which the pause between one phase ending and the next beginning is reduced or removed.¹

What does the FDA’s move to real-time clinical trials mean for sponsors and CROs?

Real-time clinical trials are not just about seeing data sooner. They create a new requirement: clinical teams must turn real-time signals into governed, human-supervised, traceable action. The FDA formalized this direction in April 2026.¹ The harder problem is execution – acting on each signal under control and that is where a governed AI Workforce fits.

Two early-phase oncology studies are leading the way: AstraZeneca’s Phase 2 TRAVERSE trial in treatment-naïve mantle cell lymphoma, and Amgen’s Phase 1b STREAM-SCLC trial in small cell lung carcinoma.¹ For the AstraZeneca study, the FDA has already received and validated real-time signals, confirming that the underlying technical framework works.¹ The agency’s real-time view is scoped by agreement: for each trial, the FDA and the sponsor set the criteria for which endpoints and signals are reported as they occur.¹ The focus is on whether regulators can act on those signals, not on streaming raw patient-level data.³

Why the FDA’s RTCT Move Matters

Early-phase trials are a particular bottleneck in drug development, marked by high uncertainty and small patient populations, with data flowing from sites to sponsors and only then to the FDA.¹ For most of clinical research history, key data signals could take years to reach regulators. “For 60 years, we’ve been conducting clinical trials in the same way, where key data signals can take years to reach the FDA,” said Commissioner Marty Makary, describing lag time that delays decisions and slows development.¹ The agency’s response is to let reviewers see safety signals and endpoints in the cloud as a trial progresses. When a patient develops a fever, or a tumor shrinks on a scan, regulators can now see it as it happens.³ The proof-of-concept trials arrived alongside a Request for Information on a broader pilot program launching in summer 2026, with selection criteria expected in July and participants finalized in August.¹ The agency followed up with an industry information session on May 15, 2026 to take questions on the proposed pilot, a sign of how actively it is moving.² Throughout, the FDA’s emphasis is on AI and data science as the enabling tools, with agency leaders pointing to a substantial opportunity to remove time from the development process and framing the effort as a way to keep U.S. biomedical research competitive.⁴ The direction is unambiguous: with FDA real-time clinical trials, clinical development is shifting from delayed, periodic review toward continuous regulatory visibility.

As continuous oversight becomes standard, execution maturity may become a key differentiator between organizations that merely observe trial risk and those that resolve it proactively.

The Real Challenge Is Not Visibility Alone

Seeing a signal earlier does not resolve it. A safety signal still needs interpretation and a documented response. An enrollment shortfall still needs a decision and an owner. A protocol deviation trend still needs escalation and follow-up. Real-time data raises the tempo of clinical operations, but it does not supply the capacity to act on everything it surfaces. This is the gap real-time data exposes rather than closes: clinical development does not lack visibility; it lacks scalable, reliable clinical trial execution capacity to resolve what visibility reveals. Faster signals without faster, governed action simply relocate the bottleneck from detection to resolution. In a real-time model, that bottleneck is no longer hidden by reporting cycles. It is on display.

What is real-time execution in clinical trials?

Real-time execution is what happens after the signal: turning a real-time data point into a governed, traceable action – a prepared query, a routed decision, a documented safety follow-up with a human in control at each checkpoint. Real-time visibility shows the change; real-time execution resolves it.

From Real-Time Data to Real-Time Execution

Real-time execution is the ability to move from signal to action with governance, traceability, and human oversight intact. It is the difference between knowing a query backlog is growing and actually clearing it; between seeing a site stall and acting in time to recover the timeline; between flagging a safety signal and closing the loop on it with a documented decision. Maxis AI was built around this distinction. Its AI Workforce for Clinical Trials operates as a supervised execution layer embedded within regulated workflows – AI agents that perform defined operational work under human validation checkpoints, controlled permissions, and clear lines of accountability. The platform’s design principle, “Designed to Think. Built to Act.”, captures the point: the value is not in observing a trial faster, but in acting on it faster, under control. This is what real-time clinical trial execution looks like in practice, and how the AI Workforce already operates inside regulated workflows today.

PROOF POINT

In a simulation-based proof-of-value study with a leading academic clinical research institute, a coordinated AI Workforce of 59 agents under expert supervision compressed roughly six to nine months of end-to-end clinical trial work into about 182 hours of agent work (approximately 7.5 days), an estimated 97% reduction in time, with human-in-the-loop review and a complete audit trail at every step.⁶

Where Sponsors and CROs Will Feel the Pressure

A continuous-visibility model concentrates pressure on the points where work actually gets done. The most exposed tend to be:

Site activation delays. Slow IRB submissions, site contracts, and activation cycles push back first-patient-in. Real-time visibility makes that delay obvious to everyone sooner, but the underlying coordination work still has to be done by someone.
Enrollment underperformance. A meaningful share of sites under-enrolls, and the shortfall often surfaces too late to course correct. Continuous visibility flags it earlier, yet someone still has to diagnose the cause and act on the right sites.
Protocol deviation trends. Recurring deviations need consistent triage, not just detection. Real-time signals make the pattern obvious quickly, which raises the expectation that it is addressed just as quickly.
Data query backlogs. Queries accumulate across EDC, labs, and eCOA faster than teams can resolve them. Watching the backlog grow in real time does not clear it; that still takes execution capacity.
Safety signal follow-up. An emerging signal demands prompt interpretation, clear ownership, and clean documentation. The faster it surfaces, the less acceptable a slow, manual follow-up becomes.
Vendor and data handoffs. Work routinely stalls in the gaps between systems and vendors. Real-time data exposes the handoff delays that batch reporting used to keep out of sight.
Study milestone risk. Small slippage in startup, enrollment, or data flow compound into timeline and budget exposure. Continuous oversight surfaces the risk earlier, but only governed execution turns that early warning into a recovered milestone.

In a real-time world, each of these becomes visible sooner which means any shortfall in the capacity to act on them becomes visible sooner too.

Why Existing Systems Are Necessary but Not Enough

EDC, CTMS, eTMF, safety, lab, and patient-facing systems remain essential. They are the systems of record, and real-time trials depend on the clean, structured data they hold. But these platforms were built to capture and display information, not to coordinate action across teams. In practice, signals still sit in disconnected systems that rarely add up to one unified operational view, and no single system assigns an owner, prepares a response, or maintains the cross-functional audit trail regulators expect. Better dashboards improve coordination, and that matters. But coordination is not throughput. Most AI systems in clinical trialsto date have improved visibility. These tools tell you what needs to happen, but they do not, on their own, increase the capacity to make it happen.

Dimension	Periodic-review model (today)	Real-time execution model (governed AI Workforce)
Data visibility	Batched; reviewed after reporting cycles	Continuous; reviewed as the study progresses
Signal-to-action	Manual, dependent on available headcount	Supervised AI execution prepares and routes the action
Oversight cadence	Periodic checkpoints	Continuous oversight with human validation
Audit trail	Assembled after the fact	Logged at every step, inspection-ready
Capacity scaling	Scales with headcount	Scales execution without proportional headcount

How the operating model shifts under real-time clinical trials.

The Role of a Governed AI Workforce

This is where a governed AI Workforce fits, the applied form of agentic AI for clinical trials. Maxis AI provides a supervised execution layer in which AI agents monitor approved inputs, reason over trial context, prepare actions, route decisions to the right people, trigger follow-ups, and maintain a complete audit trail – all within defined workflow boundaries and under expert oversight at every decision point. Built on experience from more than 3,300 clinical trials, the model expands a team’s execution capacity without proportional headcount growth, so that rising complexity and faster signals do not simply translate into more manual work and more hires. The AI agents handle structured, repeatable execution; clinical experts retain judgment, strategy, and final accountability. Examples include optimizing query management, supporting site teams with operational follow-up, and identifying emerging study risks through specialized AI Workforce agents. These execution-focused capabilities help clinical teams move from signal detection to governed action without disrupting existing systems of record.In practice, that means an agent can detect a rising query backlog, prepare and route the resolutions for a data manager to approve, and log every step so the team clears the backlog instead of watching it grow. The model works within existing systems rather than replacing them, adding the execution layer that systems of record were never designed to provide.

Governance Will Determine What Scales

The FDA has been clear that real-time visibility does not mean automated decision-making. Its Chief AI Officer noted that real-time review is not intended to replace formal sponsor-agency engagement, and that review committees themselves decide how the information is used to reach a regulatory decision.³ Legal analysts reading the initiative drew the same conclusion: the agency is signaling support for AI when it is properly governed, with documented human oversight of AI-generated outputs treated as best practice.⁵ For sponsors and CROs, that sets the bar for what will actually scale. Four controls matter most:

Role-based access

Every action is bounded by who is permitted to take it. Permissions are explicit, and the AI Workforce operates only within the workflow scope it has been granted.

Human validation checkpoints

AI agents prepare and route work, but defined decisions return to a person for review and approval. Oversight sits at every point that matters, not bolted on at the end.

Audit traceability and explainability

Every step is logged and explainable, so the path from signal to action can be reconstructed on demand, inspection-ready by design rather than reassembled under pressure.

Reproducible execution

The same workflow produces consistent, defensible output across repeated conditions. Ungoverned execution will not survive inspection; reproducible, governed execution is what compounds across a portfolio.

Maxis AI was built to that standard, with human-in-the-loop validation, audit traceability, and service-level accountability designed in from the start.

“Real-time data tells you something changed. Real-time execution is what changes the outcome – under governance, with a human in control.”

— Maxis AI

The Future: Continuous Oversight, Continuous Execution

The FDA’s longer arc points toward continuous trials, where the boundaries between phases compress and regulatory engagement is embedded throughout development.¹ That future rewards organizations that can do two things at once: maintain continuous oversight of a study, and sustain continuous, governed execution against whatever that oversight surfaces. The sponsors and CROs that move from periodic review to continuous, supervised execution will convert real-time visibility into faster, defensible decisions and ultimately into faster therapies for patients. Real-time data is the starting line. Governed execution is how teams cross it.

FAQs

What are real-time clinical trials?

Real-time clinical trials let selected endpoints, safety signals, and trial data be reviewed as a study progresses, rather than only after traditional reporting and submission cycles.

Why is the FDA moving toward real-time clinical trials?

The FDA wants to remove the lag between when trial signals occur and when regulators see them, which can delay decisions and slow development. By viewing safety signals and endpoints as atrial progress, the agency aims to accelerate promising therapies and build toward continuous trials across phases.¹

How does AI support real-time clinical trials?

AI and data science make continuous data flow and monitoring practical – detecting signals, surfacing risks, and preparing actions as data arrives. The FDA’s initiative is explicitly built on AI as the enabling technology, while keeping regulatory decisions with human reviewers.¹³

Why do real-time trials need governance?

Faster signals raise the stakes on how action is taken. In regulated environments, every step must be traceable, role-controlled, validated by humans, and reproducible. The FDA itself frames real-time review as supporting human decision-making, not replacing it.^{3, 5} Governed execution is what survives inspection and scales.

How can sponsors act on real-time trial signals?

Sponsors need an execution layer that assigns ownership, prepares responses, routes decisions to the right people, triggers follow-ups, and logs every action working across EDC, CTMS, eTMF, safety, and lab systems rather than within any single one. A governed AI Workforce provides that layer under expert supervision.

What is the role of an AI Workforce in real-time clinical trials?

A governed AI Workforce turns real-time signals into supervised, traceable action. AI agents perform defined operational work under human validation checkpoints, expanding a team’s execution capacity without proportional headcount growth – so faster signals lead to faster resolution rather than more manual load.