Artificial intelligence is transforming the world of clinical trials, promising to cut drug development timelines in half. A recent white paper from Artefact explores this ongoing revolution, highlighting innovations reshaping every stage of the process, from trial design to patient recruitment.

The paper offers insight into a dynamic ecosystem where startups, tech giants, and pharmaceutical labs are redefining the future of medical research.

Read the article on

How AI is reshaping the clinical trial landscape: an analysis by Artefact

The pharmaceutical industry is on the brink of a major revolution in clinical trials, fueled by artificial intelligence (AI). A recent white paper from Artefact, in partnership with AI for Health, explores how AI is transforming every stage of the clinical trial process, from design to result analysis.

An unprecedented opportunity to accelerate drug development

In a sector where failure is the norm – with 9 out of 10 drug candidates failing during clinical trials – AI presents a revolutionary opportunity. By reducing development timelines by several weeks, AI has the potential to save millions for pharmaceutical companies, accelerate scientific breakthroughs, and bring life-saving therapies to patients faster than ever before.

The current failure rate of clinical trials for new drugs, from phase I to final clinical approval, exceeds 90%. The main reasons for these failures include lack of clinical effectiveness (40-50%), unmanageable toxicity (30%), poor drug properties (10-15%), and absence of commercial needs or poor strategic planning (10%).

“We chose to focus on AI and generative AI in pharma R&D for several reasons. First, it’s gaining significant interest and acceleration from many stakeholders. Second, AI and R&D are crucial strategic topics today, with the potential to halve clinical trial durations. Finally, it’s a very current topic, as many laboratories are already deploying or planning to activate these use cases.”
says Thomas Filaire, supervisor of the white paper at Artefact to explains the motivations behind the study

AI is redefining each step of the clinical trial life cycle.

Artefact’s white paper explores how AI is revolutionizing the three key phases of clinical trials:

1. Clinical Trial design

AI optimizes the traditionally long and complex process of designing clinical trial protocols. Key innovations include:

  • Using language models (LLMs) to quickly extract safety and efficacy data from trial summaries.

  • Algorithmic prediction of trial success, guiding adjustments in design or focusing on more promising molecules.

  • Generating patient eligibility criteria using models like TrialGPT or AutoTrial.

These advancements significantly reduce the trial design time, sometimes cutting it from several months to just weeks. For example:

  • AI can reduce clinical protocol design timelines by an average of 30%.

  • A genetic algorithm used for pediatric bioequivalence trials reduced the number of blood sample collection points from 15 to 7, maintaining pharmacokinetic accuracy.

  • AI can shorten the time needed to analyze previous clinical trial summaries from several months to weeks.

2. Patient recruitment and inclusion

Recruiting patients remains one of the main challenges in clinical trials. AI offers innovative solutions to:

  • Predict the likelihood of patients dropping out of the trial.

  • Optimize the selection of clinical trial sites.

  • Enhance patient consent forms.

The impact of AI on recruitment is significant:

  • The time spent analyzing electronic medical records to identify eligible patients is reduced from 30 hours to 4 hours per patient.

  • A pediatric oncology study showed a 90% reduction in recruitment workload.

  • An AI tool reduced the time for selecting 90 patients to just one-fifth of the time needed manually.

  • A study in lung cancer demonstrated an AI system reviewing 102 patients in just 15.5 seconds with 91.6% accuracy for eligibility.

3. Execution and management of clinical trials

AI is transforming clinical trial execution and analysis through several advancements:

  • Risk-based quality management (RBQM): AI enhances real-time anomaly detection, enabling faster intervention and better data integrity.

  • Decentralized clinical trials (DCT): Improving participant engagement.

  • Discovery of insights via automated data analysis.

These innovations drive efficiency, reduce costs, and streamline the entire clinical trial process, speeding up drug development and improving outcomes.

  • A reduction in clinical trial costs by up to 25%

  • Acceleration of drug development by 1-2 years.

  • Shortened report writing time, from 100 to 48 days in some instances.

These innovations are reshaping the clinical trial landscape, making them faster, more efficient, and cost-effective.

An expanding innovation ecosystem.

The white paper highlights the pivotal role played by various stakeholders in advancing AI solutions for clinical trials, with a special focus on innovative startups and leading tech giants.

The leading tech giants driving innovation.

Tech giants like Google, Microsoft, IBM, and Apple are playing an increasingly vital role in advancing AI-driven clinical trials:

  • Microsoft is collaborating with Novartis on its AI for Health initiative to identify optimal trial designs and predict outcomes. This partnership focuses on using generative AI to enhance clinical trial design and improve real-time result predictions.

  • IBM Watson Health analyzes medical literature and trial data to improve protocol design and patient recruitment. Their generative AI models focus on large datasets to identify patterns in clinical data, enhancing trial protocols and recruitment strategies.

  • Verily (an Alphabet subsidiary) focuses on patient recruitment and trial management with its Baseline platform. This platform uses AI to collect and analyze health data from volunteers, allowing the design of more intelligent and personalized clinical trials.

  • Apple has received FDA approval to use its Apple Watch’s atrial fibrillation monitoring tool in clinical trials. This marks the first digital tool qualified under the Medical Device Development Tools (MDDT) program, signaling the potential to reduce costs and improve patient engagement in clinical research.

“We are moving from a reactive healthcare ecosystem to a proactive, almost predictive one.”
emphasizes Shweta Maniar, Global Director of Health and Life Sciences at Google Cloud

This shift reflects the transformative impact of AI on the entire clinical trial process.

Startups, driving innovation in clinical research.

Many startups are emerging in the clinical research field, offering innovative solutions for trial design, patient recruitment, and data management. The white paper outlines a map of these innovative players, organized according to the three key phases of clinical trials:

1. Clinical Trial design:

  • Florence Healthcare, Protocols.io, Protrials.ai: Automating protocols.

  • Perceiv AI, Seq’one: Optimizing trial targets.

  • Owkin, Insilico Medicine: Predicting trial outcomes.

  • Unlearn.ai: Creating “digital twins” of patients to predict trial results, reducing the number of control patients required by 20% to 50%.

2. Patient recruitment and inclusion:

  • Klineo, SubjectWell: Clinical trial enrollment.

  • Deep 6 AI, TrialSpark: Patient recruitment. These startups use AI to analyze electronic medical records and quickly identify eligible patients for clinical trials.

“Less than 5% of patients benefit from oncology clinical trials, while 70% say they would be willing to participate if given the opportunity. There’s a clear need for better matching between patients and trials, and recent AI advancements make this possible.”
highlights as a critical issue Thomas Peyresblanques, co-founder and CEO of Klineo

3. Execution and management of trials:

  • Lynxcare: data management

  • AiCure, Castor: decentralized trials and patient monitoring. AiCure uses AI to track patient adherence to treatments and provides real-time feedback to trial sites.

“With only 4% of trials including a representative population, Inato helps sponsors recruit patients twice as fast, increasing diversity to 67% non-white participants, compared to a previous average of 15%.”
explains Kourosh Davarpanah, co-founder and CEO of Inato, highlighting the impact of their solution

This reflects the potential of AI to improve recruitment and inclusion in clinical trials.

Challenges to overcome

Despite the promising advancements, challenges remain regarding the adoption of AI in clinical trials:

  • Regulation and Data Protection: Health authorities must adapt regulatory frameworks to incorporate new AI technologies while ensuring patient safety. Charlotte Pouchy, CEO of DEEMEA, points out, “European regulation is strict, but it can serve as an advantage by acting as a barrier to entry for non-compliant companies.” Europe is positioning itself as a key regulator, emphasizing citizen safety in AI usage.

  • Access to Data and Interoperability: Brice Miranda, Group Chief Data Officer at Servier, emphasizes: “Global data sharing is crucial to advancing research, but it is often hindered by legal, confidentiality, and competition barriers, as well as the lack of adherence to common data standards and formats.” This challenge is particularly important for rare diseases, where limited data makes collaboration even more essential.

  • Bias and Generalization: ​​AI models must be trained on representative datasets to avoid biases and ensure the results are generalizable to diverse populations. This poses a major ethical challenge, as biases in training data can lead to skewed and non-reproducible results. Addressing this challenge is critical to ensuring AI-driven clinical trials provide accurate and equitable outcomes for all patient groups.

  • Transparency and explainability: The “black-box” nature of many AI models raises ethical and trust issues, particularly in clinical trials where prediction accuracy is crucial. It is essential to explain how AI models make decisions to gain stakeholder trust. Without transparency, AI’s use in clinical trials may be limited, as regulatory bodies and patients need to understand the rationale behind decisions to ensure reliability and confidence in the technology.

  • Cybersecurity and data privacy: Data protection remains a major concern in the use of AI for clinical trials. Robust cybersecurity measures and multi-factor authentication are necessary to ensure the security of patient data.

In conclusion, AI offers unprecedented opportunities to revolutionize clinical trials, but widespread adoption requires addressing these complex challenges. Collaboration between startups, tech giants, pharmaceutical companies, and regulators will be essential to fully harness AI’s potential while ensuring the safety and ethics of clinical trials.

(1) : Zhang, B., Zhang, L., Chen, Q. et al. Harnessing artificial intelligence to improve clinical trial design. Commun Med 3, 191 (2023). https://doi.org/10.1038/s43856-023-00425-3

(2) : Tsuchiwata S, Tsuji Y. Computational design of clinical trials using a combination of simulation and the genetic algorithm. CPT Pharmacometrics Syst Pharmacol. 2023 Apr;12(4):522-531. doi: 10.1002/psp4.12944. Epub 2023 Mar 5. PMID: 36793239; PMCID: PMC10088085.

(4) : Ismail A, Al-Zoubi T, El Naqa I, Saeed H. The role of artificial intelligence in hastening time to recruitment in clinical trials. BJR Open. 2023 May 16;5(1):20220023. doi: 10.1259/bjro.20220023. PMID: 37953865; PMCID: PMC10636341.

Read the article on

.