I like about this product is its Semantic Approach to Data Loss Prevention, traditional DLP tools are incredibly rigid, they look for specific patterns, for example 16 digits of Credit Card No. Or an exact regular expression match, if any of our employees copy chunks of a proprietary source code, rephrase a confidential internal strategy into plain English, or ask public LLM to summarize these raw notes from our private Roadmap meeting, traditional security tools are completely blind to it.
June 1, 2026
It doesn't work if the user turns on private mode in the browser, and cannot detect anything running locally like electron wrappers that still contact web-based APIs. This leaves gaping holes in their ability to accurately detect all usage on user devices. Detection rules are false positive heavy and very hard to tune. Detections are not context aware, so false negatives are also high. The natural language detection engine shows promise but is impossible to tune. Support is slow to respond and not very helpful. When new updates break previous functionality, their support generally responds with links to documentation that does not help or a general
April 30, 2026