The data protection capabilities of SentinelARGUS, designed for cyberattacks in the AI era, are described, including a whitelist-based mechanism that prohibits reads from users and processes other than those authorized, as well as its scope of application.
Digital Information Technologies Corporation (DIT), taking into account the risk of generative AI misuse, describes a data protection mechanism that protects critical data after an intrusion in addition to measures designed to prevent intrusions. SentinelARGUS operates at the OS driver layer, blocking unauthorized reading, writing, encryption and deletion, and is said to protect protected data from unauthorized reads.
In addition to measures against encryption, tampering and deletion of protected data, SentinelARGUS is said to prevent information leakage by blocking unauthorized reads. The product went on sale in 2025 and is reported to control data operations based on authorization conditions.
Its main features are said to include blocking at the OS driver layer and making unauthorized access to protected files immutable. The product is also said to ensure operational management security through a two-system architecture separating the Agent and Manager.
To improve operational convenience, the product can reportedly be configured simply by selecting the protected data through a browser-based management screen, reducing implementation effort without requiring complex rule updates. It is also emphasized that legitimate business processes continue to operate as before.
Analysis of attack scenarios, based on the sequence of events at a major e-commerce operator, indicates that data theft and destruction can be contained by controlling unauthorized reading, encryption and deletion even after an intrusion. The product does not replace perimeter defenses and is intended to be used together with measures such as identity management.