MPCH and Socura collaborate to keep AI agents under human control

icon币界网
Share
AI summary iconSummary
MPCH has partnered with UK cybersecurity firm Socura to test its AI authorization platform, ATLAS, in real-world security operations. The goal is to prevent AI agents from performing critical actions—such as isolating devices or modifying system settings—without human approval. ATLAS employs cryptographic controls to define AI boundaries and features a “terminate switch” to halt unauthorized behavior. Cold signatures and offline key storage further enhance security. This collaboration combines MPCH’s expertise in crypto security with Socura’s operational experience. CEO Miles Parry emphasized the importance of integrating security early in AI and crypto applications, while Andy Kays cautioned against AI overreach. The pilot aims to mitigate potential security risks in high-stakes environments.
CoinDesk reports:

New York, November 5 /PRNewswire/ — MPCH today announced it will collaborate with UK-based Managed Detection and Response (MDR) provider Socura to test MPCH’s AI-powered access control platform, ATLAS, within Socura’s security operations.

Socura will use ATLAS verification, ensuring that AI agents for threat detection and response cannot take actions—such as isolating devices, blocking users, or modifying system settings—without authorization from a human operator.

MPCH stated that this collaboration combines its expertise in cryptographic security with Socura’s practical experience in cybersecurity operations to evaluate ATLAS’s ability to address emerging security challenges. MPCH noted that, through its defense subsidiary Ti22, the company has a verifiable track record in protecting critical operations and supporting the U.S. Department of Defense and government agencies worldwide. MPCH believes these experiences enable it to understand and deliver solutions that meet the evolving demands of increasingly data-driven and distributed operational environments.

MPCH states that traditional cybersecurity has primarily focused on protecting systems and identifying malicious activity, but AI agents have evolved from answering questions to taking actions, introducing a new challenge: ensuring no system can act without authorization. ATLAS addresses this gap by enforcing authorization before any action is executed, helping organizations prevent unauthorized actions while maintaining human control at every level as AI adoption accelerates.

MPCH Introduction: ATLAS is a control layer that applies cryptographic authorization to agentive AI, enabling operators to define the boundaries of actions an AI system is permitted to perform. Its features include a “kill switch” that can halt execution if AI actions exceed these boundaries. ATLAS also supports cold signature approval, where keys are stored in offline, isolated network-secured cryptographic storage, ensuring that no AI system or remote attacker can autonomously authorize the most sensitive operations. Human-defined permissions remain central throughout the process, while verifiable records provide auditability and accountability.

Miles Parry, Founder and CEO of MPCH, said: “AI is transforming how organizations operate, but security and authorization cannot be an afterthought. Partnering with an experienced cybersecurity team like Socura allows us to test ATLAS in real-world operations and build a solution that organizations can confidently rely on. Our goal is to ensure we’re solving actual security challenges, not just theoretical ones.”

Andy Kays, CEO of Socura, said: “While supporting clients across multiple industries, we have witnessed firsthand the challenges organizations face when integrating AI into their operations. AI agents pose a unique risk because they may act beyond their intended permissions, exposing sensitive data or disrupting critical operations.”

Jamie Brummell, Founder and CTO of Socura, added, "Maintaining human control over these actions is crucial. We believe ATLAS has significant potential to help organizations enforce these boundaries, and we are excited to collaborate with MPCH to explore how this approach can support secure and accountable AI adoption."

This partnership marks a significant step for ATLAS in deploying into high-demand environments and provides MPCH with a practical use case for applying the platform to evolving security needs driven by AI.

About MPCH

MPCH is a trusted third-party cybersecurity infrastructure provider offering high-assurance cryptographic key management, operational continuity, and disaster recovery services for institutional tokenization, enterprise, and mission-critical environments. The MPCH platform integrates cold and hot key management, HSM integration, policy-based controls, and resilient offline-capable architecture, built on a zero-trust cryptography design. MPCH’s defense subsidiary, Ti22, develops technologies that enable governments and institutions to adopt AI without relinquishing operational control or exposing their operations, restoring human control and privacy over funds, agents, and data. Together, MPCH and Ti22 provide a cryptographic foundation for global institutions and national security infrastructure. For more information, visit www.mpch.com.

About Socura

Socura is a managed detection and response (MDR) provider dedicated to bringing peace of mind and control to organizations facing an evolving threat landscape. Its 24/7/365 Security Operations Center (SOC) detects, investigates, and contains threats on behalf of clients, resolving the majority of incidents without requiring escalation to internal teams. Socura is trusted by private enterprises, public sector agencies, and critical national infrastructure, combining mature technology with expert human-led services. Socura’s SOC is CREST-certified, a signatory of the CREST AI Charter, and certified under ISO 27001 and ISO 9001. The company is ranked among the top 100 managed security service providers globally. For more information, visit www.socura.co.uk.

Disclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information. Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.