18 Nov 2026, Wed

Can Fully Homomorphic Encryption on Public Blockchains Protect Sensitive Medical Consultations? A Security and Privacy Assessment

Fully Homomorphic Encryption on Public Blockchains Protect Sensitive Medical Consultations? A Security and Privacy Assessment

Digital health services have long promised convenience and accessibility, bridging geographical divides and allowing patients to seek medical advice from the comfort of their own homes. However, this digital transformation carries a profound hidden cost regarding user privacy. For individuals seeking care for stigmatised conditions—such as mental health struggles, reproductive healthcare, or rare diseases—the fear of exposure often acts as a severe deterrent, leading many to delay or entirely avoid crucial medical treatment. While patients trust telehealth applications to safeguard their most intimate disclosures, these digital health platforms routinely expose sensitive user data and operational metadata to third parties, corporate databases, and potential malicious actors.

In an effort to address these deeply rooted vulnerabilities, researchers and cryptographers have increasingly turned their attention to emerging cryptographic primitives. Among the most promising of these is Fully Homomorphic Encryption (FHE), an advanced mathematical approach that allows computation to be performed directly on encrypted data without ever decrypting it first. Recently, FHE has unlocked new possibilities for public blockchains, enabling smart contracts to hold encrypted data whose eventual decryption is strictly governed by pre-programmed contract rules.

A new research paper published on the Cryptology ePrint Archive, authored by Nizamuddeen Shahruddeen from The University of Law, investigates whether this combination of Fully Homomorphic Encryption and public blockchains can successfully protect sensitive medical consultations. Titled “Can Fully Homomorphic Encryption on Public Blockchains Protect Sensitive Medical Consultations? A Security and Privacy Assessment,” the paper provides a rigorous examination of the promises, limitations, and subtle security pitfalls inherent in deploying FHE-enabled smart contracts for confidential healthcare communications.

To test the viability of this cryptographic defense in a real-world environment, Shahruddeen developed a functional prototype known as ConfidentialCare. Built on top of Zama’s fhEVM—an innovative framework that brings confidential smart contracts to the Ethereum ecosystem—and deployed on the Ethereum Sepolia testnet, the prototype was designed to simulate a private telehealth interaction. In the ConfidentialCare architecture, a patient’s symptoms and medical queries are encrypted directly within their web browser before ever touching the blockchain. Once secured on-chain, only a specifically authorised doctor possesses the cryptographic capability to decrypt the consultation messages.

The author evaluated the security posture of ConfidentialCare through a multifaceted methodology. The evaluation framework incorporated rigorous metadata inspection, comprehensive access-control tests, and targeted attack experiments. Throughout this assessment, the research sought to answer two fundamental questions: what level of data protection does FHE actually afford, and can access to sensitive patient information be reliably and securely controlled within a decentralised blockchain environment?

The evaluation yielded nuanced findings, dividing the privacy protections into clear categories of success and vulnerability. On the positive side, the actual content of the medical consultations remained entirely confidential across all tested scenarios. Because the data was encrypted via FHE prior to transmission and processing, any observer monitoring the public ledger could not read the medical symptoms, diagnoses, or advice exchanged between the patient and the physician. Furthermore, the design of ConfidentialCare successfully eliminated the collection of traditional personally identifiable information, requiring no names, email addresses, or conventional account registrations.

However, the security assessment uncovered significant privacy leaks at the metadata layer, highlighting inherent limitations in public blockchain architectures. Even though the message contents were securely hidden by Fully Homomorphic Encryption, public on-chain metadata laid bare critical patterns of communication. Specifically, anyone inspecting the public ledger could determine which blockchain wallet consulted which doctor, identify the medical specialty of the treating physician, record the exact timestamps of the consultations, and deduce the approximate length of each transmitted message. The research notes that if a user’s wallet address were ever deanonymized or linked back to their real-world identity in the future, this entire historical record of healthcare interactions would be permanently exposed.

Beyond metadata leakage, the security assessment revealed critical vulnerabilities regarding access control and data re-sharing. The study discovered that a doctor who was legitimately granted access to a patient’s encrypted data could subsequently re-share that information with unauthorized parties. This alarming capability manifested through two distinct vectors: a confused-deputy flaw embedded within the application’s smart contract logic, as well as a direct capability inherent in the underlying fhEVM access control list. Furthermore, testing revealed that neither formally withdrawing access nor officially closing a consultation removed the doctor’s decryption privileges under the initial configuration.

Recognizing these vulnerabilities, the author engineered a robust architectural redesign. By implementing a system based on time-limited, revocable delegation, the revised framework successfully prevented doctors from arbitrarily re-sharing sensitive patient data. Zama’s live service was able to enforce revocation and expiration policies immediately. Testing on the Sepolia testnet demonstrated that Zama’s key-management service firmly refused decryption requests that lacked a valid, active delegation. Notably, the revocation of access took effect within the exact same block across all six experimental trials, proving the cryptographic mechanism’s operational reliability.

Ultimately, the security and privacy assessment concludes that while Fully Homomorphic Encryption successfully shields the actual disclosures made by patients, it fails to obscure the broader context and mere fact of seeking medical care on a public ledger. Moreover, the research highlights a critical operational nuance for developers working with privacy-focused blockchain frameworks: within fhEVM, the technical permission granted to a user to decrypt data is functionally equivalent to a permission to share it, necessitating careful, time-bound delegation protocols to maintain true confidentiality in sensitive domains like healthcare.

Leave a Reply

Your email address will not be published. Required fields are marked *