IBM has released proof-of-concept technology designed to protect sensitive data used in industries such as healthcare without the need to modify existing applications.
Masking Gateway for Enterprises (Magen), unveiled on Thursday, is designed to catch protected data before it reaches users' screens, IBM said. The technology, which runs on a server installed alongside existing server and client applications, was developed at IBM's Haifa research laboratory and takes its name from the Hebrew word for 'protection' or 'shield'.
Magen treats a screen of information as a picture and uses optical character recognition to identify the parts that an administrator has identified as confidential, IBM said. The software then places a mask over the details that need to remain hidden, without ever copying, changing or processing the data itself.
It is designed for cases when information needs to be shared either internally or externally, for example to outsourcing organisations or marketers.
"Magen provides a common solution for all applications, regardless of their operating system and communication protocols," said Tamar Domany, project leader at IBM's Haifa lab, in a statement. "The solution is completely generic and can be used with any data, any application, and for different levels of authorisation."
Existing data-masking technologies generally make copies of the data with the protected elements masked, according to IBM, but this process of copying creates more data that needs to be kept secure. Magen is designed to eliminate the need to make such copies by masking protected information on the fly.
"Using optical character recognition technology and predefined configurations enables us to provide masking without changing the client applications," Domany said.
The system can be quickly adapted by the administrator to respond to the introduction of new privacy regulations or new types of users, IBM said. The company has applied for patents on the word-scrambling and image-manipulation techniques used in Magen.
Magen competes with data-masking software from rival companies such as DataGuise and Oracle.
