One patent tells you about an invention. The assignee map — who holds how many patents in a field — tells you about a market. Homomorphic encryption is a good place to read that map, because the field is large enough to have a shape and recent enough that the shape is still forming.
A search of the homomorphic-encryption space (CPC H04L 9/008, the subclass for the technique) returns thousands of granted records, and the assignee distribution is concentrated at the top among names you would expect from an enterprise IP race large technology and financial firms with deep portfolios, alongside a long tail of specialists. That breadth is itself the headline homomorphic encryption is no longer a handful of academic groups; it is a field where many large enterprises are actively staking ground.
The representative grants this desk has read map onto that distribution. Microsoft's US10541805B2 (variable relinearization, 2020) sits in the efficiency layer; Samsung's US11575502B2 (homomorphic processing device, 2023) is a hardware play; Enveil's US10880275B2 (secure analytics, 2020) is an application-layer stake from a focused firm. Different assignee types — platform giant, hardware maker, specialist — claiming different layers of the same field is the classic structure of a maturing technology market.
Underneath all of it sits the academic root, and the contrast is the strategic point. The foundational bootstrapping claim, Stanford's US8515058B1, is a university grant. The commercial cluster is built on the field that academic work opened. When the deep foundational position is academic and the broad commercial layer is corporate, freedom-to-operate analysis gets interesting — the enterprises are competing on improvements, not on the core.
The trend direction matters as much as the snapshot. The year-over-year count of homomorphic-encryption grants has climbed steadily through the late 2010s and into the 2020s, with recent years among the most active. Rising filing velocity in a CPC subclass is the portfolio-level tell that an area is heating up commercially — companies file when they expect the field to matter to their products and their competitors'.
The caveats this column always attaches assignee counts and filing velocity are signals of investment and intent, not proof of product success or technical leadership. A firm can hold many patents and ship nothing; another can hold few and dominate. But the map is real evidence of where effort and money are going, and for homomorphic encryption the map says the same thing the individual claims do a research idea has become an enterprise contest, built on an academic foundation, growing year over year.