Deep Instinct has announced the general availability of DIANNA (Deep Instinct’s Artificial Neural Network Assistant), a new tool designed to provide enhanced explainability for unknown cybersecurity threats by working with its DSX Brain deep learning framework.
DIANNA operates in conjunction with the DSX Brain, which is described as the first purpose-built deep learning (DL) cybersecurity “brain.” The introduction of DIANNA addresses the cybersecurity industry’s historical reliance on an “assume breach” mindset, which has contributed to significant financial losses from attacks. Legacy detection-and-response tools are often characterized as slow and reactive, struggling against AI-powered threats. IBM data indicates that organizations typically take an average of 241 days to identify and contain a breach.
Lane Bess, CEO of Deep Instinct, stated, “Just as ChatGPT 5 delivers clear answers to complex questions, DIANNA brings instant threat explainability to cybersecurity teams. DIANNA delivers the clarity and confidence needed to explain unknown attacks prevented by the DSX Brain, accelerating the work of security teams by enabling faster, more decisive action. This is the future of cybersecurity: prevention-first, intelligent, and transparent.”
The DSX Brain is central to Deep Instinct’s prevention-first strategy. It is a proprietary discriminative neural network developed specifically for cybersecurity, having been trained on tens of billions of data points over the past decade. The DSX Brain is designed to continuously learn, stopping known and unknown threats in under 20 milliseconds, with over 99% accuracy and less than 0.1% false positives.
While the DSX Brain focuses on stopping threats before execution, DIANNA provides the contextual understanding of these events. Utilizing generative AI, DIANNA functions as a virtual team of malware analysts, translating complex, previously unseen threats into actionable, easy-to-understand narratives. In under 10 seconds, DIANNA delivers insights into attack anatomy, patterns, and behaviors for both known and unknown threats, providing security teams with the necessary context for more effective and faster responses.
Deep Instinct notes that DIANNA differs from basic bots offered by competitors, which often rely on known threat data from existing logs or reputation data. DIANNA is designed to explain previously unseen attacks, offering analysts expert-level visibility into unknown threats. This capability aims to increase visibility and confidence in the DSX Brain’s verdicts, while also contributing to a reduction in false positives and the streamlining of Security Operations Center (SOC) workflows.
Through the combined capabilities of the DSX Brain and DIANNA, Deep Instinct aims to shift the cybersecurity industry from reactive detection to preemptive data security. Deep Instinct states it is the first company built on a deep learning cybersecurity framework to prevent unknown threats in under 20 milliseconds. The Deep Instinct Data Security X (DSX) platform secures data at-rest or in-motion across cloud environments, Network Attached Storage (NAS), applications, and endpoints. Information on DIANNA’s capabilities is available at the DIANNA Explains Hub. A webinar demonstrating the technology is scheduled for Wednesday, October 1st at 11 AM ET.
