Highlights from SCOPE 2026
Acceleration and agentic AI were the key words at this year’s Summit for Clinical Operations Executives (SCOPE) congress (held 2–5 February in Orlando, Florida, USA). PSI was well-represented, with Kasia Moscicka, Head of Feasibility and Site ID, co-presenting with Guillaume Bartoletti from Sanofi on the importance of collaboration in site ID, and Emily McInturf, Feasibility Director, exploring the interplay between AI and human interaction in feasibility. Read on for our key takeaways from this important conference to understand the paradigm shift in our industry towards smarter, AI-driven support to accelerate clinical trials.
We are at a pivotal point in AI: The push for faster, more cost-effective trials and improved patient and site experiences, coupled with sponsor’s desires for almost instantaneous, actionable insights is driving the shift towards AI for many traditional roles (e.g., developing statistical analysis plans or data management plans). Orlando was the meeting point for more than 4500 clinical research professionals, representing over 1000 companies from 30 countries, to discuss the solutions to the needs of clinical trials of the future – here are our highlights for sponsors from this enlightening meeting.
Speeding up clinical trial execution
AI-driven strategic velocity
The overall timeline from molecule discovery to market authorization has remained largely unchanged over the past 20 years [1], despite innovations such as EDC, centralized monitoring, smart SDV, and decentralized trial models. These tools have increased productivity within segments of the clinical trial timeline, but the gains have not resulted in overall end-to-end timeline compression. While the advent of new modalities (gene therapies, personalized medicine) is partially responsible for the unchanging timelines, attendees at SCOPE believe AI can change this.
Advances in data infrastructure and AI introduce a new capability – systematizing and scaling decision logic across studies and portfolios. This enables:
- Earlier decisions
Real-time integration of historical/operational/external data to identify risks, optimize site selection, and refine protocols reducing uncertainty prior to starting - A greater degree of parallel activities
Traditionally sequential tasks (e.g., feasibility, site ID, operational planning) can be performed concurrently without compromising quality – through development of standardized decision frameworks and predictive analytics
However, acceleration efforts face some significant hurdles:
- Wide variability in site capabilities and workflows complicates standardization
- Regulatory and quality requirements slow the pace of innovation adoption
- Technology adds administrative burden to sites, reducing time for patient-focused activities
- Human interaction is essential and cannot be automated
Where is opportunity the greatest?
Leaders across industry expect AI to deliver 30%–40% reduction in clinical development timelines within the next 2–5 years, and that clinical trial workflows will not be reactive, linear, or manual, but autonomous, integrated, predictable, multi-step processes. With the AI in clinical trials market projected to reach $8 billion by 2030, the focus is currently on developing AI-driven early-stage activities (late-stage acceleration is likely to progress more slowly):
- Protocol design
- Site ID and selection
- Startup
- Site enrollment readiness
Sponsors can expect to define their execution strategy much earlier in the trial development lifecycle. Front-loading the analytical effort and decision making will have a great impact on enrollment speed, amendment rates, and overall cost and trial duration.
A new approach to site ID
Kasia Moscicka, Head of Feasibility and Site ID at PSI, and Guillaume Bartoletti, Manager of Clinical Trial Operations at Sanofi, co-presented on the importance of close collaboration – and how established processes and tools allowed them to identify 80% of needed sites in just six weeks for a four-study phase 3 program enrolling across 40+ countries. Their talk also unveiled PSI’s new platform, SYNETICTM, a one-click approach to site ID using agentic AI. SYNETIC combines 30 years of therapeutic expertise and operational data, with insights from more than 500,000 institutions and 3,000,000 sites in a semantic knowledge platform to improve trial efficiency and accelerate timelines.
To be able to identify the vast majority of sites for Sanofi’s important trial in such a short time relied upon mutual respect and trust – and understanding their needs and the intricacies of the trial. Site ID is 80% relationships, and 20% enablement – with SYNETIC, we now have more time to focus on the relationships pillar.
New approach to site ID
Adding a semantic knowledge platform to solid process foundation.
Results
>80% of target in less than 6 weeks
Agentic AI and redesigning workflows
Perceptions on AI
AI is advancing at a rapid rate and is reshaping our expectations of the future. Approximately 88% of companies are integrating AI into their operations, and are mostly split among three camps:
- The early adopters
The larger pharma companies and some CROs are the most proactive in developing and adopting AI into their world. They lead the way in embedding AI into their operations, from trial design and site ID to enrollment strategies and data analysis. Many hold meetings where sites are able to share their thoughts and feedback on the new AI innovations reaching them. Several pharma companies and CROs were able to demonstrate early results at SCOPE, highlighting measurable gains in trial acceleration (and treatment efficacy as a by-product) - The pragmatic users
Small-to-mid sized companies generally adopt a pragmatic, reactive response to AI. Rather than regarding it as a transformative agent to redesign workflows, they adopt targeted tools that solve specific operational challenges. AI is considered an efficiency enhancer, rather than a collaborative assistant - The latecomers
For some, regulatory concerns, limited expertise, data governance challenges, or other uncertainties means they are slow or hesitant to adopt AI. Waiting for clearer standards, or proof of value, means they may reduce risk, but also will widen their operational gap as the early adopters race ahead with accelerating timelines
It is crucial to consider AI, in particular agentic AI, as a central component in clinical trial design and operations – and not simply as bolt-on tools to address specific issues. Sponsors partnering with CROs experienced in AI application can be confident that their trials will be as efficient and effective as possible.
From potential to practice: Unlocking AI workflows
At SCOPE, the consensus was that a major challenge lies in reengineering clinical processes (i.e., not just digitizing them). For example:
- Future protocols will evolve from static PDFs to dynamic documents, updated in real-time, that can proactively facilitate improvements and reduce clinical trial complexity (and increase speed)
- A first step is to simulate patient pathways and experiences, and therefore identify potential obstacles before initiating the protocol
- Trials in indications or areas with historically low enrollment rates could be , assessing protocol/site/patient burden and pinpointing the challenges to solve
- For example, natural language processing could scan the protocol, different algorithms can be developed to compare against historical databases/health records, and recruitment speed could be predicted based on site performance patterns and disease prevalence metrics
- Data collection and review is a bottleneck in the workflow. Strategic SDV is a start (driven by AI-powered insights on which data requires a closer look), however a future where data is collected, stored, and cleaned instantaneously could again accelerate clinical trials, and improve time to market
- In practice, this could take the form of integrated data storage and transfer, combined with natural language processing of clinical notes and AI mapping and flagging of source data fields (with machine learning to detect unusual patterns or outliers, not just rule violations)
Implementing AI tools means that more time is made for teams to cultivate better relationships with PIs, provide site training and support, and to focus on innovation and implementation. AI is about reducing manual work – to take on operational tasks and allow human intelligence (HI) the time needed for considered strategic decision-making.
Using AI to plan studies with confidence
Emily McInturf, Director of Feasibility at PSI, presented the results of a collaboration between PSI and clients (co-developed with ITM) on using AI, feasibility and financial data, and HI to plan studies and portfolios. PSI’s VISIONALTM uses a blend of historical study data, industry insights, and costing data to model studies and portfolios in real time. Full budgets and cash flow projections are also delivered for each scenario. This approach helps to build program and trial strategies in minutes, tailored to specific criteria/goals, preferred regions, known country startup timelines, milestones, and budget constraints.
VISIONALTM helps sponsors understand the feasibility and financial impact throughout the portfolio upfront. Being able to have the important conversations on what matters most take place as soon as possible means we sponsors can control the timelines and costs from the very start, accelerating their trials.
Assess risk tolerance
Use case: Portfolio-level modeling
Effective adoption of AI
New outsourcing models
Accelerating trials through use of AI begins with selecting the right tools for the right challenges. Sponsors must balance efficiency, cost, and the true problem-solving capability of each solution.
Selecting an AI tool requires the same level of oversight and operational partnership as selecting a CRO. The right tool(s) can accelerate your study – but the wrong tool(s) may break it.
Questions that should be asked before an AI purchase could include:
- Is longitudinal hands-on training and support provided?
- Is there a human ultimately responsible for project supervision as part of the service?
- Is the AI ready for use, or will it be a custom-build project?
- Is the AI validated and certified, per relevant industry standards?
- How can we mitigate the risks with using AI?
- Are benchmarks and validation data available for our specific use case?
AI is currently being used full-circle across the clinical trials industry – for example, AI used to request RFPs/RFIs, to develop them, and ultimately to score them for vendor selection. Whilst time-saving (increasing time-to-award), mitigation of risk should be implemented (i.e., an operations team with a strong clinical background to interpret AI-generated results, and a vendor team with significant AI experience). Ultimately, effective use of AI should free up humans to focus on making informed, strategic decisions and reducing the repetitive manual work slowing innovation.
Developments in risk-based quality management
RBQM remains a core focus, nearly a decade after its introduction in ICH GCP R2. Many sponsors are still refining how to translate risk assessment into meaningful operational efficiencies. SCOPE highlighted that as R3 expectations evolve, sponsors will need stronger RBQM oversight across CRO partnerships, reinforcing the need for structured, risk-driven approaches.
Effective AI adoption represents a natural progression of RBQM. The value lies in not generating more data, but prioritizing what is critical to quality. Attendees at SCOPE raised important questions, such as whether queries can be prevented upstream rather than resolved later, how to apply the 80/20 principle to prioritize meaningful data, and whether all data collected requires active cleaning or intervention.
Effective RBQM adoption is moving towards predictive, agentic AI-enabled models. Agentic AI can autonomously plan and interpret complex datasets (including unstructured data) and identify emerging risks with actionable recommendations to safeguard trial quality and timelines.
Key takeaways
SCOPE’s central theme of accelerating clinical trials was woven throughout all discussions, however the industry was reminded that speed alone is not the objective – acceleration matters because it brings therapies to patients faster. AI is not technology for technology’s sake – it is about improving outcomes for patients.
AI fundamentally changes how we think about clinical trials. As it continues to mature, human intelligence remains central. Agentic AI is capable of planning, coordinating, and acting toward complex goals, and is beginning to enter overall operational workflows and increase trial velocity. AI enables parallel decision-making, digestion of large quantities of data, and proactive risk and suggested action. The next major breakthroughs will come in unlocking AI in daily workflows and in integration of outsourcing models.
At the same time, challenges remain about data quality (and quantity), governance, validation, and regulatory expectations, reinforcing that human expertise and oversight are still essential to guide decision-making and ensure responsible AI adoption.
Our challenge is to focus on the tangible applications that deliver value and acceleration today, while remaining open to an exciting future where AI and human expertise continue to evolve together.
Explore PSI’s AI-driven clinical trial solutions
Abbreviations
AI, artificial intelligence;
CRO, Contract Research Organization;
EDC, electronic data capture;
HI, human intelligence;
ICH GCP, International Council for Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use Guidelines for Good Clinical Practice;
ID, identification;
ITM, Isotope Technologies Munich;
RBQM, risk-based quality management;
RFI, Request for Information;
RFP, Request for Proposal;
SDV, source data verification.
References
1. Brown, D.G., et al., Clinical development times for innovative drugs. Nat Rev Drug Discov, 2022. 21(11): p. 793–794.
