The AI chip industry sector has attracted enormous amounts of capital for high performance computing with a heavy focus on cybersecurity. This recent Wall Street Journal article highlights almost $10bn of venture investments in the last year.
If the cybersecurity space is of interest to you, then you need to be paying attention to a new technology out there: “fully homomorphic encryption”. Fully homomorphic encryption (or FHE) allows sensitive data to be processed and analysed, all while remaining fully encrypted. This ensures reliability and security while the data passes into the hands of second and third parties. FHE has countless real world applications. Check out the compelling overview on FHE included in the sites below:
Take this hypothetical, for example – imagine a medical researcher is trying to compute descriptive statistics on a population of lung cancer patients at a hospital. Unfortunately, the hospital is unable to share its private medical records with the researcher without violating HIPAA. With FHE, the hospital can encrypt its data using a fully-homomorphic scheme. This allows the data to remain protected and anonymous, but also computable.
In this particular case, the hospital would encrypt its medical records and send them to the medical researcher’s cloud computing environment. Because the data is encrypted, it is protected and fully private in the cloud. Next, the researcher runs its analytical functions on the homomorphically-encrypted data in the cloud, manipulating the data while it remains encrypted. Last, the researcher downloads the encrypted output and decrypts the result to reveal the plaintext answer. Notice that the sensitive medical record data is encrypted end-to-end and is only decrypted when revealing the final answer behind organisational firewalls.
By understanding the practical implications of FHE, it is easy to understand why investors are so excited about this new technology. Here at Impact Venture Capital, we are optimistic about FHE for several reasons.
First, it is impossible to completely avoid third parties when communicating or storing sensitive data. FHE means users no longer need to trust third parties to keep their word on privacy because the data is always encrypted, secure, and private even in untrusted environments.
Second, FHE eliminates the tradeoff between data usability and data privacy – there is no longer any need to mask or drop any features to preserve the privacy of data. All features may be used without compromising privacy.
Finally, the number of billion dollar markets requiring an FHE solution is growing by the month – banking, healthcare, government, insurance, and other regulated data-centric global industries all require more robust cybersecurity solutions. Can we stop criminals from laundering money through the complicated secure banking system? Can we fend off the next worldwide pandemic by understanding private patient healthcare information? Can we help protect sensitive data in a war time situation? Can we improve information exchange of sensitive data between business units and vendors? Can we process and validate private information included in elections to ensure accuracy of citizen’s votes? FHE will likely be a part of the solution in all of these cases.
Although some of the above markets and use cases may seem complicated, there are also a large number of obvious markets and simple use cases. Telecom companies, retailers, mail order companies, public sector organisations, credit card companies, and others hold a large amount of privacy relevant and highly confidential data. Many get hacked and pay with huge reputation loss and regulatory fines. Class action lawsuits often follow. If data would be encrypted at the source (retail shopping and e-commerce) and always kept encrypted throughout the transaction, the risk of being hacked is much lower or would completely disappear. Building a reputation as a company that securely holds customer data could be a huge differentiator these days.
As IBM and Intel increase their focus on FHE and the cybersecurity market this year, one of our portfolio companies, Cornami, is viewed by many as the market leader in the FHE space. Cornami is a high-performance computing company that develops breakthrough computing architecture for real-time environments. It empowers developers, large enterprise, IoT, and edge-to-cloud computing to deliver high performance anywhere and on any device at the lowest power and latency. This is achieved by efficiently using a high volume of small cores in a highly concurrent, parallel manner.
In 2021, Cornami partnered with Inpher, a pioneer in secret computing, to collaborate on delivering FHE functionality to the marketplace. Existing FHE algorithms are computationally intensive and have been often considered as not yet practical for real world applications. Cornami’s partnership with Inpher overcomes such limitations to deliver real-time FHE computing to a ready and rapidly expanding market. Here is a white paper developed by the Cornami team and made accessible through the global tech forum for semiconductor professionals
Over the past few months, Cornami closed on more than $65m of funding with a lead investment from SoftBank (the largest venture investor in the world) including co-investments from Applied Materials (NASDAQ: AMAT) and Impact Venture Capital. Here is the recent industry media coverage by the high-profile technology publication.
FHE might not bring the glamour of the metaverse but it will transform the world.
First published on LinkedIn