As data breaches and cyberattacks become increasingly sophisticated, traditional encryption methods face unprecedented threats.

The rise of quantum computing also poses a significant risk to current encryption methods, which could be rendered obsolete by the computational power of quantum. Additionally, the exponential growth in machine learning and artificial intelligence heightens the need for secure computing, as these technologies rely heavily on vast and high quality datasets.

If compromised data is fed into AI models, the resulting outputs will also be compromised, therefore, ensuring the quality, integrity and accuracy of data, in addition to its volume, is critical. Fully Homomorphic Encryption (FHE) offers a way forward, poised to transform how we handle and share sensitive data.

The data dilemma: protecting vs utilising 

Data is often referred to as the most valuable global asset. However, its true value is only realised when used to make informed decisions – be it improving operational efficiency, developing products or understanding societal trends. Organisations are increasingly seeking ways to optimise this value through new technologies such as AI, ML and data collaboration. However, valuable data often remains siloed within organisations and the most valuable data is usually the most sensitive.

Data breaches by criminal organisations can also have devastating consequences, not only for the organisation, but for the individuals whose personal data has been stolen. This data must be kept confidential and shared only with trusted parties. However, the need for collaboration introduces tension between the benefits of data sharing and the risks to confidentiality.

Encryption is typically applied to sensitive data only when data is being moved or stored. To process data, it typically needs to be decrypted first, exposing it to risks. This presents a dilemma – protecting data and limiting its use or utilising the data and increasing exposure to breaches.

FHE resolves this tension by enabling encrypted data to be computationally processed. Data can be shared without ever being exposed or vulnerable, making it useless to attackers even if intercepted. FHE is ushering in a new era of secure computing and supporting the new data economy by allowing multiple parties to work on the data without ever actually accessing it.

Challenges of FHE and the potential of silicon photonics

Despite its immense potential, FHE has faced significant adoption challenges, primarily due to its substantial computing power requirements and the inefficiencies of traditional electronic processing systems. FHE requires specialist hardware and considerable amounts of processing power, leading to high energy consumption and increased costs. However, FHE enabled by silicon photonics — using light to transmit data — offers a solution that could make FHE more scalable and efficient.

Current electronic hardware solutions systems are reaching their limits, struggling to handle the large volumes of data and meet the demands of FHE. However, silicon photonics can significantly enhance data processing speed and efficiency, reduce energy consumption and lead to large-scale implementation of FHE. This can unlock numerous possibilities for data privacy across various sectors, including healthcare, finance and government, in areas such as AI, data collaboration and blockchain. This could potentially lead to significant progress in medical research, fraud detection and enable large scale collaboration across industries and geographies.

The path to widespread adoption

The Covid-19 pandemic highlighted the real-world outcomes when organisations collaborate effectively for a shared goal. Vaccine development, typically a lengthy process, was accelerated through big pharma companies working together. For example, the partnership between BioNTech, Fosun Pharma, and Pfizer led to the rapid development of the widely distributed Pfizer-BioNTech vaccine. This involved sharing large amounts of unique and valuable information, including biomedical data and trial results – often without formal agreements in the early stages. However, this also highlighted the risk of compromising sensitive information and the need for better tools to ensure data security and confidentiality.

Privacy Enhancing Technologies (PETs) have traditionally been complex and challenging to deploy. However, FHE stands out by its ability to maintain full cryptographic security, which ensures data remains protected against unauthorised access during processing. This allows data scientists and developers to run data analysis tools on sensitive information without ever seeing or compromising sensitive data. While implementing FHE presents challenges for users without cryptographic skills, modern FHE software tools are making it increasingly accessible without requiring deep cryptographic knowledge. Additionally, regulatory environments are evolving to support widespread FHE adoption. Guidance from bodies like the Information Commissioner’s Office (ICO) and regulatory sandboxes in regions like Singapore are supporting the development of FHE. Its applications are vast, spanning government-level data protection, cross-border financial crime prevention, defence intelligence exchange, healthcare collaboration, and AI integration.

In healthcare, for example, FHE can enable secure analysis of patient data, supporting advanced research while ensuring patient data remains confidential. Financial institutions can perform secure computations on encrypted data for risk assessments, fraud detection, and personalised financial services. Government and defence companies can also enhance national security with secure communication and data processing in untrusted environments. Additionally, FHE allows for the secure training of machine learning models on encrypted data, combining AI’s power with data privacy.

The future of data security with FHE enabled by silicon photonics

FHE is set to transform the future of secure computing and data security. By enabling computations on encrypted data, FHE offers new levels of protection for sensitive information, addressing critical challenges in privacy, cloud security, regulatory compliance, and data sharing. While technical challenges remain, advancements in FHE technology are paving the way for its widespread adoption.

As we continue to generate and rely on large amounts of sensitive data to solve some of society’s biggest challenges, FHE enabled by silicon photonics provides a secure and efficient solution that ensures data can be used and remain confidential. The future of secure computing is one where organisations can do more with their data, either through secure sharing or processing — unlocking its full potential without compromising privacy.

Nick New is the CEO and founder of Optalysys.

Read more

Why data isn’t the answer to everything – Splunk’s James Hodge explains the problem with using data (and AI) in helping you make key business decisions

Data encryption: what can enterprises learn from consumer tech? – Siamak Nazari, CEO of Nebulon, discusses the data encryption lessons that enterprises can learn from consumer tech