SuperCom Unveils SafeGenie for Real-Time Monitoring and Control of Autonomous AI Agents

Delivers novel, patent-pending technology to address the security risks of agentic AI, with initial rollout planned through the global customer base and distribution channels of Safend, SuperCom’s cybersecurity subsidiary

Combines new AI-agent controls with Safend’s endpoint security technology, developed over more than two decades, deployed across millions of endpoints and backed by 27 issued patents

SuperCom, a global provider of secure solutions for the IoT, cybersecurity, and digital identity sectors, introduced SafeGenie, a cybersecurity product suite for the real-time monitoring and control of autonomous artificial intelligence (AI) agents. SafeGenie combines novel, patent-pending methods for agentic AI security with the kernel-level endpoint security technology of Safend, SuperCom’s wholly owned cybersecurity subsidiary. Safend’s technology has been deployed across millions of endpoints for thousands of customers, including government agencies, defense organizations and enterprises.

AI agents are spreading rapidly across enterprises, yet traditional security was built around people, devices and approved applications, not software that acts with full user permissions and decides its own next step in a statistical and sometimes unpredictable way. Security budgets are beginning to reflect this shift. Gartner forecasts that spending on securing AI will grow from approximately $2.8 billion in 2026 to almost $7.7 billion by 2028,¹ which represents average growth of approximately 65% a year.

SafeGenie assesses the context and intent behind each action an AI agent takes and allows, flags or blocks it in real time, before the action completes, helping prevent agent misbehavior, sensitive-data leakage and irreversible damage. The suite also secures conventional generative AI use, including prompt protection, as part of a comprehensive control plane for AI across the enterprise.

Also Read: CIO Influence Interview with John Elliott, Cybersecurity Author Fellow at Pluralsight

“Agentic AI presents enterprises with a major opportunity. Agents can now research, move data and complete multi-step tasks across business systems with a degree of independence that makes them exceptionally productive,” commented Ordan Trabelsi, President and CEO of SuperCom. “That same independence makes every agent a potential insider threat, operating with an employee’s access, at machine speed and around the clock. Our customers should not have to choose between productivity and control. SafeGenie is designed to give them that control, stopping high-risk actions while legitimate work carries on.”

“Our cybersecurity engineers have developed novel, patent-pending methods to address security challenges unique to agentic AI,” Trabelsi added. “Kernel-level security is difficult to build and stabilize at scale, and Safend’s technology has been proven across thousands of customer environments. We believe that foundation, combined with these innovations and our long-standing relationships with government, defense and enterprise customers, positions us well to secure agentic AI for both our existing and new customers.”

Claude Code, OpenAI Codex, GitHub Copilot and Claude Cowork are among the most widely used agentic AI tools in the enterprise, alongside shadow AI: agents such as OpenClaw installed without IT approval, unapproved large language models (LLMs) and agents running under personal or out-of-domain accounts or stolen identities. Acting with an employee’s permissions or even broader access, and faster than any person can review, an agent can reach data it was never meant to use, leak sensitive information, follow instructions hidden in an email or file, or break out of its boundaries. The risk is no longer theoretical: in July 2026, AI agents under evaluation at OpenAI escaped their test environment and breached production systems at Hugging Face, an incident both companies have publicly disclosed.

Rather than relying on network traffic inspection alone, SafeGenie addresses these risks where agents operate, including on the device itself, from the kernel level beneath the applications and agents it governs. Each decision follows the organization’s own policy rather than an AI model’s judgment. Because it also operates locally, SafeGenie can protect supported endpoints wherever they are, online or offline, including locally run agentic AI.

The initial rollout will focus on integrating SafeGenie with Safend’s infrastructure and offering it to Safend’s existing customers, directly and through Safend’s distribution channels across the United States, Europe and Asia. SuperCom then plans to offer SafeGenie to its government customers and additional enterprises. SafeGenie will be offered as a separate subscription, in addition to existing Safend licenses and maintenance agreements.