Humanising Artificial Intelligence In Biopharma and Healthcare
Using blockchain as information gatekeeper
Frank Leu, Ph.D., Founding member, BioPharMatrix LLC
There is an urgent need for humanising AI in biopharma and healthcare. This can be accomplished using blockchain to fortify data transparency, security, and patient control. Blockchain will be the data gatekeeper to ensure the production of high-quality datasets for training AI in order to facilitate personalized drug research and development. Using smart contracts and encryption would then support ethical, patient-centered AI within the proper regulatory frameworks.
The integration of artificial intelligence (AI) into biopharmaceutical and healthcare holds immense promise for transforming drug discovery, clinical trials, and personalized medicine. However, the success of AI models hinges on the quality, security, and transparency of the data they are trained on. Ultimately, when applying automated processes to human health, we must ensure that the highest ethical standards are infused throughout the datasets that AI is trained on to create a functional “humanised AI.” Blockchain technology, with its inherent characteristics of immutability, transparency, and decentralization, presents a unique opportunity to address these data challenges and pave the way for standardizing humanised AI. This article discusses how a well-designed blockchain system can enhance biopharma innovation by providing a secure and transparent data infrastructure for AI training, ultimately leading to more personalized and patient-centric pharmaceutical and healthcare solutions. In the following sections, we explore practical implementations of blockchain in biopharma and how they contribute to ethically grounded, trustworthy AI outcomes.
Blockchain and AI in biopharma and healthcare
The biopharmaceutical and healthcare industries are beginning to recognize the power of blockchain to address data-related challenges. Blockchain can secure patient data, improve supply chain transparency, and facilitate secure data sharing for research purposes (Swan, 2015). For instance, it can create a tamper-proof audit trail for clinical trial data, ensuring data integrity and compliance with regulatory requirements. In practice, such benefits are already being realised. In 2022, Mayo Clinic began using a blockchain-integrated platform for a multi-center clinical trial, creating an immutable audit trail of trial data while supporting remote patient monitoring and electronic consent (Fox, 2022). This allowed data from over 500 patients across 10 sites to remain tamper-proof and accessible to all trial stakeholders in real time, thereby upholding integrity and patient engagement in the study (Fox, 2022). Similarly, major pharmaceutical companies have explored blockchain in the drug supply chain. A notable FDA-supported pilot, the MediLedger project, brought together manufacturers, distributors, and pharmacies to track prescription drugs using a shared ledger (Fulton, 2020). By registering drug shipments as unique digital tokens on a blockchain, the system could automatically enforce that only one authorized party owns a given drug package at any time and flag any inconsistencies, greatly improving traceability and helping to prevent counterfeit medications (Fulton, 2020). Beyond R&D and supply chains, blockchain has also been implemented to secure sensitive patient records. For example, Estonia’s national e-health system uses blockchain technology to safeguard the health data of its 1.3 million residents, logging every access to patient records in an immutable ledger (Einaste, 2018). This approach ensures that medical data cannot be tampered with and that any use of patient information is transparently recorded — a foundation of trust critical for AI-driven healthcare services. Each of these cases illustrates how a reliable, transparent data infrastructure can augment AI: When AI algorithms train on data that is verifiably secure and unaltered, the resulting insights are more trustworthy and ethically grounded.
AI, on the other hand, excels at analyzing large datasets to identify patterns and insights that can accelerate drug discovery and personalized treatment (Topol, 2019). However, AI’s reliance on data quality raises concerns about bias and lack of transparency. The combination of blockchain and AI addresses these issues by providing a secure and transparent data foundation for AI training. To this end, studies have shown that blockchain can enhance data provenance and traceability, which are crucial for building trust in AI-driven healthcare solutions (Dagher et al., 2018). By leveraging blockchain’s immutability, AI models can be trained on verifiable datasets, reducing the risk of bias and improving the reliability of their predictions. This leads to the concept of “humanised AI,” where the AI can better understand and simulate human-centric outcomes through validated training data. In other words, an AI system whose training data has been “filtered” or vetted by blockchain’s gatekeeping is more likely to produce outputs that are transparent, trustworthy, and aligned with human values.
Standards for the gatekeeper blockchain in training humanised AI
To achieve humanised AI in biopharma and healthcare, blockchain systems should be designed as gatekeepers of data. At a minimum, they should meet a few specific criteria (outlined below) to infuse data with proper ethical and quality standards:
1) Apply smart contracts to automate data sharing and ensure fair data usage. Smart contracts, which are selfexecuting code on the blockchain, can be programmed to automatically enforce data-sharing agreements and consent policies. For example, a smart contract might allow an AI algorithm to access anonymized patient data only after the patient’s digital consent has been recorded on the blockchain and then log that access as a transaction. Such contracts ensure data is used according to predefined rules without requiring manual oversight. In supply chain settings, smart contracts have been used to enforce business rules (i.e., preventing double ownership of a drug shipment), illustrating their power to maintain data integrity and fairness in real time (Fulton, 2020).
2) Embed transparency and auditability to create an auditable trail of all data transactions. Every interaction with data should be recorded on the blockchain ledger, creating a transparent, immutable log. This means that at any point, authorised stakeholders (including auditors or even patients) can trace how data was added, accessed, or modified. Such an audit trail directly addresses the “black box” issue in AI by providing external accountability for the data going into and out of algorithms. Notably, in 2017, Google’s DeepMind Health unit proposed a Verifiable Data Audit ledger to record every access of patient data by its AI systems, precisely to bolster transparency and public trust in how AI was using sensitive health information (Vincent, 2017). On a blockchain, similar audit logs are inherently built in, and they cannot be altered after the fact — providing confidence that the AI’s inputs (and even outputs) can be reviewed and trusted.
3) Develop standardized protocols to allow interoperability among stakeholders. Standardisation is key to sharing data seamlessly between different healthcare participants. A blockchain intended to unite data from pharma companies, clinics, and patients must speak a “common language” with existing systems. This could involve adopting healthcare data standards, like Fast Healthcare Interoperability Resources (FHIR) standard developed by Health Level Seven International (HL7), for electronic health records or the GS1 standard by the Electronic Product Code Information Services (EPCIS) for supply chain events. and ensuring the blockchain network can interface with them. In the MediLedger pilot, for instance, the solution was designed to be interoperable with other systems and standards, demonstrating how a blockchain network can integrate into the broader healthcare data ecosystem (Fulton, 2020). By using common protocols, data from a hospital’s database, a clinical trial management system, and a pharmaceutical inventory can all be shared on one ledger, enabling AI models to draw insights from a much richer, more diverse pool of information than any siloed system would allow.
4) Ensure data integrity and security through robust encryption and access controls. Sensitive scientific and medical data must be protected both on-chain and off-chain. Blockchain’s immutability secures integrity (no record can be altered surreptitiously), but confidentiality is achieved through encryption. In practice, raw patient data might remain off-chain in secure databases, while only cryptographic hashes or pointers to the data are stored on-chain (Yaga et al., 2019). Access to the actual data can then be restricted to those holding the proper decryption keys or permissions. Permissioned blockchains (where only vetted nodes can join) further enhance security by limiting who can even participate in the network. Moreover, advanced cryptographic techniques can augment privacy. For example, zero-knowledge proofs allow one party to prove certain data or statements are true without revealing the underlying data itself — a method that was incorporated in the MediLedger design to validate transactions without exposing confidential business information (Fulton, 2020). By combining these measures — encryption, permissions, and cryptographic privacy techniques — blockchain systems ensure that even while data is being shared and audited, personal or sensitive information remains secure from unauthorized eyes.
5) Execute patient-centricity by empowering patients to control their data and participate in research. A truly humanised AI requires that patients have control over their information. Blockchain can enable decentralised identity and consent management, wherein patients use digital keys or identities to directly grant or revoke access to their health data. For instance, a patient could use a blockchain-based portal to consent to their genomic data being used in an AI-driven cancer research project, and that consent with its scope and duration would be immutably recorded. Patients could just as easily withdraw consent, with the blockchain logging the revocation and smart contracts automatically preventing any further data access. Such dynamic consent management has been demonstrated by solutions like the one developed by Bitfury and Longenesis, where participants in medical studies can withdraw consent at any time, and the change is recorded in real time on the ledger (Garrity, 2019). Empowering patients also means potentially rewarding them for data contributions — an idea emerging in some blockchain health platforms. For example, Triall’s clinical trial platform is tokenized, allowing patients or participants to be compensated (via cryptocurrency tokens) for their data or involvement (Fox, 2022). This not only incentivizes patient participation but also aligns with ethical data practices by recognizing patients’ contributions. By putting patients in control — deciding who can use their data, tracking how it’s used, and even benefiting from its use — blockchain provides a mechanism to keep AI development anchored to patient rights and needs.
By adhering to the above criteria, a blockchain system can create and fortify a trusted data ecosystem that supports the development of a humanised AI capable of delivering personalized and effective healthcare solutions. In essence, the blockchain serves as a vigilant gatekeeper: only data that meets integrity, consent, and quality checks make it through to the AI algorithms, and every step is logged. This gatekeeping approach means AI models learn from data that is not only high-quality but also ethically sourced and transparently handled — ultimately making the AI’s outcomes more human-aligned and acceptable to end-users.
Synergistic Clinical Impact on Patient Outcomes
The synergy between blockchain and AI can lead to significant improvements in clinical outcomes for patients. Personalized treatment strategies can be derived from AI models trained on blockchain-secured patient data — models that can identify tailored therapies based on individual genetic profiles and comprehensive medical histories. In oncology, for example, an AI system might comb through a blockchain-backed database of cancer patients’ genomic data and treatment responses to identify patterns that predict which therapy would be most effective for a new patient. Because the underlying data is verifiable and comes with patient consent, clinicians and regulators can have greater confidence in these AI-driven recommendations, accelerating the adoption of genuinely individualized treatments.
Blockchain-enabled data sharing can also facilitate collaborative research and accelerate the identification of new drug targets, thus speeding up drug discovery and development. Researchers from around the world could securely contribute and access a shared pool of clinical and molecular data on a blockchain network. AI algorithms analyzing this rich, diverse dataset might uncover, say, a novel protein target implicated in a rare disease. Since each dataset (perhaps contributed by different hospitals or companies) is traceable and its integrity is assured, follow-up experiments can proceed more rapidly, building on a trusted foundation. Indeed, industry analysts suggest that combining advanced AI analytics with a decentralized data framework can dramatically increase the success rate of clinical trials by improving data integrity and participant engagement (IBM, 2021).
Blockchain can streamline clinical trial processes by ensuring data integrity and facilitating secure, real-time data sharing among researchers, sponsors, and regulatory agencies. Key trial documents and results written to a blockchain are immediately available to authorized parties and cannot be altered, reducing delays and errors in data reconciliation. For instance, patient-reported outcomes or adverse events could be submitted via a blockchain-based system; AI tools could monitor these incoming data streams for safety signals or efficacy trends. Suppose an AI detects an issue (perhaps an unexpected side effect pattern). In that case, investigators can trust that the underlying data points are genuine (given the tamper-proof ledger) and take swift action. Automating such data monitoring with AI, while relying on blockchain for trustworthy data, could lead to improved clinical trial efficiency and safety oversight. Ultimately, transparent and secure data management builds and enhances trust between patients, healthcare providers, and researchers. Patients enrolled in a trial might feel more at ease knowing they can see how their data is being used and that it’s protected, potentially improving enrollment and retention. Healthcare providers and regulators, on the other hand, gain confidence in trial outcomes, knowing that the evidence collected is auditable and has not been manipulated.
The more comprehensive and well-structured the dataset generated by a robust blockchain, the more nuanced and “humanised” the AI algorithms can become, reflecting a deeper understanding of human biological and behavioral complexities. As data from various sources (clinical, genomic, lifestyle, etc.) are securely integrated, AI can build multidimensional models of health and disease. These models won’t be opaque amalgamations of miscellaneous data; rather, they’ll be constructed from datasets with clear provenance and consent. In practice, this means an AI might not only predict which treatment a patient will respond to, but also provide an explanation grounded in the patient’s verified data profile — bridging the gap between complex analytics and human interpretability.
Humanised AI built for an efficient trusted biopharma and healthcare ecosystem
The integration of blockchain and AI has the potential to revolutionize healthcare by creating a trusted and patient-centric ecosystem. However, several challenges must be addressed to realize this vision fully, including regulatory compliance, data privacy, and operational interoperability:
1) Regulatory compliance and data privacy: Any solution deploying blockchain and AI in healthcare must navigate a complex web of laws and regulations. In the United States, for example, the Health Insurance Portability and Accountability Act (HIPAA) mandates strict standards for protecting patient health information. In the European Union, the General Data Protection Regulation (GDPR) imposes requirements for data consent, transparency, and the right to be forgotten. A blockchain-based system needs to be designed in a way that can uphold these rights and protections. This is not trivial, given that blockchain’s immutability means data written to the ledger cannot be easily erased. There is an apparent tension between GDPR’s requirement that personal data be erasable upon request and the blockchain principle that ledgers are append-only and permanent. Early implementations are finding ways to reconcile this — for instance, by keeping personal data off-chain and only storing references or encrypted tokens on-chain, which can be rendered unusable (effectively “forgotten”) if needed. Solutions like the Longenesis consent platform have demonstrated that it’s possible to build HIPAA-compliant and GDPR-compliant healthcare applications on blockchain by careful architecture choices (Garrity, 2019). Still, the industry will need to establish clear guidelines on how to handle data deletion requests, consent withdrawal, and other patient rights in the context of an immutable ledger. Regulators are beginning to engage in these questions. The FDA’s exploration of blockchain for drug tracking and the EMA’s interest in AI transparency are positive signs that authorities are open to evolving regulatory frameworks. Ongoing dialogue between technologists and policymakers will be essential to adjust regulations (or the technology) in a way that balances innovation with privacy and safety.
2) Operational interoperability and collaboration: Beyond legal considerations, there are practical challenges to integrating blockchain and AI into existing healthcare workflows. Hospitals and research centers often use legacy electronic record systems that were not designed to connect with distributed ledgers. Ensuring interoperability means not only agreeing on data standards (as discussed earlier) but also upgrading infrastructure and training personnel. Collaboration between healthcare providers, researchers, and technology developers is essential to overcome these barriers (Topol, 2019). Stakeholders must agree on governance: who operates the blockchain network, who has access to what data, and how are disputes resolved? Consortium blockchains in healthcare require trust and cooperation among competitors — an area where clear value propositions (such as fighting counterfeit drugs or sharing AI insights for rare diseases) can motivate participation. Initiatives like industry consortia or public-private partnerships can help here. For example, the MediLedger consortium showed that even fierce competitors in pharma could work together when the goal was a secure, shared supply chain system. Similar collaborative models might be needed to build blockchain-based data networks for AI in clinical research, where multiple hospitals or companies pool data for the common good. Interoperability also extends to the AI side: models and tools need to be able to consume blockchain-sourced data. This might involve developing APIs or middleware that let AI systems query the blockchain or subscribe to real-time data feeds from it. Efforts are underway in standards organizations and groups like IEEE and ISO to define best practices for blockchain in healthcare, which will facilitate smoother integration across systems.
3) Ethical and transparency considerations: To fully realise the potential of humanised AI, we must ensure ethical and responsible use of blockchain and AI in healthcare. This means establishing norms for algorithmic transparency — patients and clinicians should have some insight into how AI recommendations are generated. Blockchain can assist by providing an immutable record of the data and even the algorithm versions used at each point in time, which could be reviewed in case an AI decision is questioned. There is also the question of accountability: if an AI makes an erroneous recommendation that leads to harm, the presence of a clear data audit trail would help in investigating and understanding the root cause (whether it was faulty data, a biased algorithm, or misuse of the recommendation). Patient empowerment, as discussed, is both an ethical imperative and a contributor to better outcomes. When patients are treated as partners in the data ecosystem, their trust in AI-driven interventions is likely to increase. We must strive to avoid scenarios where patients feel their data is exploited by opaque algorithms. Instead, the goal is a transparent ecosystem where patients know their data rights and all AI actions are traceable. Achieving this will likely involve new standards, for example, an ethical blockchain-AI charter or certification and perhaps oversight bodies that audit AI systems using tools like blockchain logs.
In sum, marrying blockchain with AI in biopharma and healthcare requires more than just technology; it demands a supportive framework of policies, standards, and cross-sector collaboration. Encouragingly, initial steps are being taken, regulatory sandboxes, pilot programs, and standards development are all in progress in different parts of the world. These efforts need to continue and expand.
Concluding thoughts
Blockchain technology offers a powerful solution for enhancing data security, transparency, and patient-centricity in the biopharmaceutical industry. By providing a trusted data infrastructure for AI training, blockchain serves as the gatekeeper of selective and critical information that can be used to humanize AI models and deliver personalized, effective healthcare solutions. The convergence of these technologies, as illustrated through examples in clinical trials, supply chain management, and health data sharing, indicates a transformative potential for healthcare delivery. Future research and development should focus on refining standardized protocols that are consistent with both internal policies and governmental regulations to foster wider collaboration among stakeholders. Emerging standards and regulatory guidance (from HIPAA and GDPR to FDA and EMA initiatives) will play a pivotal role in shaping implementations, and adhering to them will build further trust in the system.
Ultimately, by continuing to invest in blockchain as an information gatekeeper, we can accelerate the development of AI that is not only intelligent but also ethical, transparent, and truly centered on the human aspects of healthcare. The journey to humanised AI in biopharma is just beginning, but with sustained collaboration and innovation, its profound benefits—from more effective treatments to a more trustworthy healthcare ecosystem—are within reach. The stakeholders in biopharma and healthcare — from researchers, patients, service and material providers to pharma companies and regulators — must work together to address the challenges and ensure that the deployment of these powerful tools remains aligned with the core values of biopharma and healthcare: do no harm, patient first, and advance scientific knowledge for societal benefit.
Author Bio
Frank Leu, Ph.D., is a pharmacologist and the founding member of BioPharMatrix LLC, dedicated to advancing disruptive solutions in biopharma and life sciences. Frank has extensive research experience in molecular biology, cancer biology, and metabolic disease. Frank has over a hundred presentations and article publications combined, contributing to scientific research and biopharmaceutical innovations.
