Sevii Unleashes Autonomous AI Warriors for Identity Cyber Defense

CHARLOTTE, N.C. – February 17, 2026 – As cyber adversaries increasingly weaponize artificial intelligence to launch attacks at machine speed, cybersecurity firm Sevii has responded with a new force of its own: “Agentic AI Cyber Warriors.” The company today announced the general availability of its Autonomous Identity Security module, a system designed to fight fire with fire by deploying fully autonomous AI agents to defend the most targeted part of a modern enterprise—digital identity.

The new module is part of Sevii's broader Level 5 Autonomous Defense & Remediation (ADR) platform, which aims to detect, investigate, and neutralize threats without requiring human intervention. This move comes as identity-based attacks, where hackers use compromised credentials to breach networks, have surged by 68% year-over-year. While detection tools can raise an alarm in minutes, overwhelmed security teams often take hours or days to respond, creating a critical window for attackers to cause significant damage.

The New Battlefield: Identity Under Siege

Digital identity has become the soft underbelly of corporate security and the primary entry point for cyberattacks. According to a recent report from Palo Alto Networks' Unit 42, nearly two-thirds of all initial network intrusions in the past year involved identity-based techniques. The problem is pervasive, with industry data from 2024 showing that 90% of organizations experienced at least one identity-related incident, and 84% of those breaches resulted in a direct business impact.

Attackers are no longer just manually searching for weak passwords. They are leveraging dark AI and large language models to scale their operations, crafting sophisticated phishing campaigns and using stolen credentials harvested from previous breaches. IBM's X-Force team reported that identity-based attacks constituted 30% of all intrusions they responded to in 2024, a trend largely fueled by the automation of credential harvesting. This creates a severe mismatch where automated attacks moving at machine speed are met with manual, human-paced defenses.

“Adversaries Aren't Waiting for Humans to React — Why Should Defense?” asked Curt Aubley, CEO and Co-Founder of Sevii, in the company’s announcement. “In today's cyber battlespace, attackers are using autonomous AI to identify and exploit identity weaknesses in minutes. Defense must run at the same speed, not at the pace of analyst queues and manual triage.”

Rise of the Machines: Level 5 Autonomous Defense Explained

Sevii's answer to this challenge is what it calls “Level 5 Autonomous Defense.” The term, borrowed from the autonomous vehicle industry, represents a state of full automation where a system can operate entirely without human oversight under all conditions. In cybersecurity, this translates to AI agents that can independently process threat alerts, hunt for related malicious activity, analyze the context, and execute a response—such as disabling a compromised account—all in a matter of minutes.

While the concept of a fully autonomous, Level 5 security system is still an aspirational goal for the industry at large, it represents a clear direction away from tools that simply create more work for analysts. The progression is seen on a spectrum, from Level 0 (fully manual) to Level 5, where each level introduces more sophisticated automation. Sevii's platform aims to push the industry toward the higher end of this spectrum, where AI agents become an extension of the security team, actively performing the work of investigation and remediation.

Sevii claims its approach yields dramatic results. Early customer deployments of the Autonomous Identity Security module have reportedly demonstrated over 95% improvements in key security metrics: Mean Time to Detect (MTTD), Mean Time to Hunt (MTTH), and Mean Time to Remediate (MTTRR). A reduction of this magnitude would be transformative, shrinking response timelines from days or hours down to mere minutes and drastically reducing the dwell time an adversary has inside a network.

Trust and Control in the Age of AI Warriors

The prospect of AI operating “without humans in the loop” raises critical questions about governance, trust, and control. How can an organization be sure an autonomous agent won't mistakenly lock out a CEO or shut down a critical system? Sevii emphasizes that its platform is built with strong AI governance to address these concerns. “Sevii invented Autonomous Defense & Remediation and our new Autonomous Identity Security module addresses the speed and scale of identity-driven compromise... with the AI governance customers expect and demand,” stated Aubley.

Robust AI governance in cybersecurity requires several layers of safeguards. These include transparency through explainable AI (XAI) so that actions can be audited, strict operational guardrails to limit the scope of autonomous actions, and the non-negotiable presence of a manual “kill switch” for human operators to override the system if necessary. As AI agents themselves become powerful actors within a network, securing their own identities is also becoming a new frontier for companies like Microsoft and Okta, which are developing solutions to manage and govern these nonhuman workers.

This reflects a fundamental shift in philosophy. “Cyber Outcomes Are What Lowers Risk, Alerts Don't,” said Stephen Collins, CTO of Sevii. “Security isn't measured by noise or analyst activity—it's measured by how quickly systems are protected. This isn't augmentation. This is AI Automation with AI Governance, and it's redefining what real defense looks like.”

A Crowded Field of AI-Powered Guardians

Sevii is entering a competitive and rapidly evolving market. Established cybersecurity giants are all heavily investing in AI to bolster their identity security offerings. CrowdStrike’s Falcon Identity Protection uses AI to protect human and non-human identities, while Okta offers AI-driven threat detection and automated responses within its Identity Security Fabric. Microsoft has integrated its Security Copilot AI assistant into its Entra identity platform, and Palo Alto Networks continues to build out its identity security capabilities.

In this crowded field, Sevii is differentiating itself with its bold focus on full autonomy and its “agentic AI” model, which promises to do the work rather than just flag it for humans. This vision has garnered significant industry recognition. Sevii was recently selected for the prestigious 2026 CrowdStrike, AWS & NVIDIA Cybersecurity Startup Accelerator, an endorsement from leaders in cloud, AI, and security. The company was also named the “Most Innovative AI Cybersecurity Product Company” for 2025 by Cyber Defense Magazine. Such accolades suggest that Sevii's pioneering approach to autonomous defense is being viewed as a credible and potentially transformative force in the fight against next-generation cyber threats.