Sevii Extends Autonomous Defense Platform with New AI Security Module for Real-Time Remediation

New AI Security Module for Autonomous Defense

Sevii has introduced a new AI security module as an extension to its Autonomous Defense & Remediation (ADR) platform. This module is specifically designed to counter the increasing speed and scope of AI-driven cyberattacks. It operates at runtime, independently of the source of shadow AI within an organization, providing immediate and autonomous remediation capabilities.

Real-Time Alert Ingestion and Analysis

The new module integrates with a customer's existing security detection stack, ingesting alerts in real-time. Unlike traditional tools that typically report detections to a Security Operations Center (SOC), Sevii's AI module intercepts these reports. It then initiates an instant, AI-driven response at machine speed to address the detected threats.

AI-Driven Attack Confirmation and Remediation

The system employs AI agents, referred to as 'cyber warriors,' to perform a seven-day retrospective context hunt. This process determines whether a detected action is normal or abnormal, thereby confirming if it is a genuine AI attack. If an attack is confirmed, these agents investigate the possibility of the same attack occurring elsewhere in the customer's infrastructure to assess its broader scope and determine the necessity of immediate remediation.

Addressing the Speed of AI Attacks

According to Sevii's CEO and co-founder, Curt Aubley, the module immediately collects and analyzes data to reverse engineer attacks and take necessary actions. While a 'human in the loop' option exists, the company emphasizes that effective defense against AI attacks requires machine-speed reactions, as human response times are insufficient to counter rapid, automated threats. Aubley cited an example of 17 seven-minute actions by rogue OpenAI agents attacking Hugging Face, highlighting the mathematical impossibility for humans to keep pace.