CyberProof Inc., a co-managed security services company owned by UST Global Inc., today launched an agentic artificial intelligence service that it says can autonomously handle up to two-thirds of security investigations while leaving human analysts in control of critical decisions.

The service, called CyberProof Agentic MXDR, is a form of managed extended detection and response that connects AI agents with human security experts across the detection, response and exposure management lifecycle. It replaces manual security operations center workflows with what the company calls a human-governed ecosystem of collaborative agents, mapping that orchestration to a client’s own environment and tying it to threat data and existing defensive tools to speed up alert triage.

CyberProof frames the launch as a step past standalone AI tools. The service pairs SOAR playbooks, the security orchestration, automation and response scripts security teams already run, with AI agents that handle alert triage, hunt for threats and tune detection rules. People still make the calls. The agents work under human governance and pass only the hardest cases up to analysts, which CyberProof said sharpens investigation accuracy by as much as 30%.

Unlike single-agent tools or closed AI SOC platforms, CyberProof said, its framework works with models from Microsoft Corp., Google LLC, Anthropic PBC and other third parties, alongside client-developed and CyberProof-built agents, without vendor lock-in. A built-in quality control framework continuously scores those agents on effectiveness, speed and cost, giving chief information security officers a read on accuracy versus compute spend as frontier model prices rise.

The pitch rests on speed. Generative AI has compressed the time attackers need to exploit new flaws and frontier models can surface thousands of zero-day vulnerabilities across major software and operating systems, CyberProof argues, leaving calendar-driven defense unable to keep pace with the rate of disclosures.

“In the last few months, we’ve seen how AI-powered threats can compress the attack lifecycle to minutes, forcing organizations to operationalize threat intelligence faster than ever before,” said Chief Executive Tony Velleca. “Collecting threat data is no longer a competitive advantage. The advantage lies in execution, reducing exposure, immediate detection and containment. Because of this speed gap, legacy security operations metrics are obsolete.”

Doron Davidson, a CyberProof vice president, said the blend of AI agents and human expertise is aimed at faster, more consistent investigations with fewer false positives.

The 24/7 service reports outcomes through CyberProof’s Reveal360 dashboard and is available immediately. The company said it’s optimized for large enterprises and mid-market organizations running complex hybrid cloud environments and is offering a complimentary readiness assessment for teams weighing their operational maturity and stack compatibility.

Founded in 2017, CyberProof delivers co-managed security operations built around Microsoft Azure and the wider Microsoft security stack.

Image: CyberProof