Company leaders are continually looking for ways to keep data safe without compromising its usability. Fully homomorphic encryption (FHE) could be a step in the right direction.
What Is Fully Homomorphic Encryption?
Fully homomorphic encryption allows the analyzing and running of processes on data without needing a decryption method. For example, if someone wanted to process information in the cloud but did not trust the provider, FHE would allow sending the encrypting data for processing without providing a decryption key.
Read more: Creating a Cloud Strategy: Tips for Success
How Does Fully Homomorphic Encryption Work?
FHE is like other encryption methods that require using a public key to encrypt the data. Only the party with the correct private key can see the information in its unencrypted state. However, FHE uses an algebraic system that allows working with data without requiring decryption first. In many cases, information is represented as integers, while multiplication and addition replace the Boolean functions used in other kinds of encryption.
FHE uses an algebraic system that allows working with data without requiring decryption first.
Researchers first proposed FHE in the 1970s, and people became interested back then. However, it has taken substantial time to turn these concepts into feasible real-world applications.
A researcher showed it was plausible with his 2009 published study. However, working with even a tiny amount of data proved too time-intensive. Even now, FHE can require hundreds of times more computing power than an equivalent plaintext data operation.
What Advantages Does FHE Have Over Other Types of Encryption?
Data is at a higher risk of becoming compromised when it’s not encrypted. FHE keeps the information secure by not requiring decryption to occur for processing to happen.
In one recent example, Google released an FHE-based tool that allows developers to work with encrypted data without revealing any personally identifiable information (PII). Google’s blog post on the subject gave the example of FHE allowing medical researchers to examine the data of people with a particular condition without providing any personal details about them.
Encryption takes private information and makes it unreadable by unauthorized third parties. However, something that makes people particularly excited about FHE is that it eliminates the tradeoff between data privacy and usability, making both present at a high level.
Is Fully Homomorphic Encryption Safe?
Many people familiar with FHE and its potential applications agree that it seems safer than other methods of data protection, which require decrypting data for processing. It could be particularly widely embraced in certain sectors. After all, cloud computing brings in $250 billion per year.
Experts believe FHE will emerge as a compelling option in tightly regulated industries.
People are continually interested in how to keep their data safe when stored in the cloud. Some experts also believe FHE will emerge as a compelling option in tightly regulated industries because it could become a better safeguard against breaches.
“Past solutions to either completely anonymize data or restrict access through stringent data use agreements have limited the utility of abundant and valuable patient data,” IBM notes on its site. “FHE in clinical research can improve the acceptance of data-sharing protocols, increase sample sizes, and accelerate learning from real-world data.”
What Are the Business Use Cases of FHE?
Fully homomorphic encryption could forever change how companies use data. That’s crucial, especially considering how many businesses collect it in vast quantities at a time where many consumers feel increasingly concerned about keeping their details safe.
For example, FHE allows keeping information in an encrypted database to make it less vulnerable to hacking — without restricting how owners can use it. That approach could limit an organization’s risk of regulatory fines due to data breaches and hacks.
It also permits secure data monetization efforts by protecting customers’ information and allowing services to process people’s information without invading privacy. In such cases, individuals may be more forthcoming about sharing their information, knowing in advance that business representatives cannot see certain private aspects of it.
Using an FHE-based solution also enables sharing data with third-party collaborators in ways that reduce threats and help the company providing the information comply with respective regulations. Thus, this kind of encryption could support research efforts where people across multiple organizations need to work with sensitive content.
Which Companies Offer FHE Products?
Fully homomorphic encryption is not widely available in commercial platforms yet. However, some companies offer products based on homomorphic encryption that could eventually work for the use cases discussed earlier.
For example, Intel has such a product that allows segmenting data into secure zones for processing. Similarly, Inpher offers a product with an FHE component. It primarily uses secure multiparty computation, but applies FHE to certain use cases.
IBM says FHE is now adequate for specific use cases.
Beyond those examples, IBM has a fully homomorphic encryption toolkit that it released for iOS in 2020. That progress primarily occurred after IBM’s experts took it upon themselves to make FHE more commercially feasible, addressing the time and computing power that it previously took to use this type of encryption.
The company’s representatives say FHE is now adequate for specific use cases and suggested the health care and finance industries as particularly well suited to it.
Fully Homomorphic Encryption Shows Potential
Since FHE is not widely available via commercial platforms yet, interested parties should not expect to start using it immediately. However, that could change as organizations become increasingly concerned about striking the right balance between data security and usability.
The ideal strategy for businesses to take now is to explore the options currently on the market. They can then determine if any of those options check the boxes for helping them explore fully homomorphic encryption, including what it might do in the future and what capabilities exist now.