Beyond Firewalls: Atsign Redefines AI Security with 'Invisible' Agents

SAN JOSE, CA – January 28, 2026 – As artificial intelligence rapidly integrates into every facet of the enterprise, a new and complex set of security challenges has emerged. In recognition of a groundbreaking approach to this problem, Atsign, a company specializing in pre-emptive security, has been named a finalist for the Most Innovative AI-Infrastructure Security Solution in SiliconANGLE Media’s 2026 Tech Innovation CUBEd Awards.

The nomination celebrates Atsign’s atPlatform, a technology that seeks to fundamentally reshape how AI systems are secured. Rather than reinforcing traditional digital perimeters, the platform is designed to make AI agents and their underlying infrastructure invisible to attackers from the outset, a concept that is gaining significant traction as businesses grapple with the vulnerabilities of a hyper-connected AI landscape.

The Tech Innovation CUBEd Awards, presented by the influential digital media leader SiliconANGLE Media, are known for spotlighting technologies and leaders shaping the future of the industry. “The finalists in this year’s Tech Innovation CUBEd Awards are some of the most ambitious and effective innovators in technology today,” said Dave Vellante, co-founder and co-CEO of SiliconANGLE Media. “Each company, product and leader recognized has demonstrated significant vision and execution in shaping the future of B2B and B2B2C technology.”

The Growing Security Debt of AI

The explosion of generative AI has created immense opportunities but has also introduced critical security risks, often categorized as 'AI Sprawl' and 'AI Security Debt.' AI Sprawl describes the burgeoning complexity of managing and securing a vast, decentralized network of specialized AI agents. As organizations deploy more models to handle various tasks, the potential attack surface expands exponentially.

Compounding this issue is AI Security Debt—the accumulation of vulnerabilities that arise when AI systems are deployed rapidly on top of legacy infrastructure that was never designed for them. This friction between cutting-edge AI and outdated security models is a primary reason many AI pilot programs fail to scale. The atPlatform was engineered to address this core problem by providing a new abstraction layer that unifies identity, encryption, and data transport.

Traditional security measures like firewalls, VPNs, and network segmentation rely on a 'protect the perimeter' model. This approach becomes increasingly ineffective when AI agents need to communicate dynamically across different cloud environments and even between separate companies. Such a model is often slow, cumbersome, and riddled with potential points of failure that sophisticated attackers can exploit.

A Paradigm Shift to 'Invisible Networking'

Atsign’s nomination stems from its radical departure from this perimeter-based philosophy. The atPlatform introduces what the company calls 'invisible networking,' a security posture built on the principle that you cannot attack what you cannot see. It achieves this by eliminating the need for open network ports, a primary vector for cyberattacks like port scanning and unauthorized access attempts.

At the heart of the platform are 'Restricted Access Agents' (RAAs). These are described as invisible, self-sovereign software entities that operate without a discoverable network footprint. Each agent is assigned a unique, cryptographically verifiable identity called an 'atSign' (e.g., @myAIagent). Instead of connecting via IP addresses and open ports, communication occurs through a peer-to-peer, end-to-end encrypted protocol managed by the atPlatform. Every connection attempt is authenticated and authorized at the edge, creating a true Zero-Trust environment where trust is never assumed.

This architecture fundamentally changes the security dynamic. By default, every device, agent, and service running on the platform is dark to the public internet. Connections are only established between two parties who have mutually authenticated each other's atSign identity. This preemptively closes the door on a wide range of common network-based attacks, shifting the focus from reacting to threats to preventing them from ever materializing.

Enabling the Federated AI Economy

The business implications of this technology are profound, particularly for the development of Federated Multi-Agent Systems (MAS). These systems, where autonomous AI agents from different organizations collaborate on complex tasks, are considered a cornerstone of the future AI economy. However, enabling them securely has been a major roadblock.

Atsign's platform is designed to dissolve this barrier. By handling identity and secure transport seamlessly, it allows development teams to spin up dynamic, load-sharing AI agents across different clouds and company boundaries without the typical delays and complexities of configuring firewalls, VPNs, or load balancers.

"We are honored to be recognized by theCUBE and SiliconANGLE for redefining how AI agents interact," said Barbara Tallent, CEO at Atsign, in the company’s announcement. "The future of the AI economy depends on federated systems where agents from different companies can collaborate safely. By shifting the paradigm from 'protecting the perimeter' to 'securing the agent’s intent,' we’ve created the sweet spot where CEO vision meets CISO approval—allowing companies to deploy autonomous AI at the speed of code, not the speed of network tickets."

This capability promises to accelerate innovation by allowing enterprises to securely leverage specialized AI models from partners or vendors without exposing their internal infrastructure or sensitive data, fostering a more collaborative and powerful AI ecosystem.

Industry Recognition and Advanced Capabilities

The recognition from SiliconANGLE Media underscores the industry's need for such forward-thinking security frameworks. John Furrier, co-founder and co-CEO of SiliconANGLE Media, noted the high caliber of the finalists. “Being a finalist in the 2026 Tech Innovation CUBEd Awards is a testament to the strength of each company’s ideas, resilience and impact,” he stated.

Beyond its core architecture, the atPlatform includes several industry-first security features that further enhance its capabilities for mission-critical AI deployments. One is a 'Kill Switch,' which allows for the dynamic and immediate revocation of an agent's access rights, even in the middle of a session. This provides a powerful tool for instantly containing a compromised agent or responding to an emergent threat.

Another innovative feature is the 'Blind Visibility' model. This allows an organization to enforce security and data policies on agent interactions while remaining cryptographically blinded to the sensitive content being exchanged. The system can verify compliance without ever needing to decrypt the data, ensuring that privacy is maintained while governance is upheld. As AI continues its relentless march forward, solutions that build security and privacy in by design, rather than treating them as an afterthought, will be critical for realizing its full potential safely and responsibly.