en.Wedoany.com Reported - Dubai, UAE-based OTT Cybersecurity LLC officially released the Agent Trust Protocol (ATP) on May 11, 2026. This is the first open cryptographic standard for AI agent identity, permissions, and action verification, and it is planned to be submitted to the Internet Engineering Task Force (IETF) to become an industry-wide specification. According to the company's official announcement, ATP covers five cryptographic primitives—identity, scope, attestation, delegation, and revocation—allowing any system to verify in real-time whether a communicating entity is an authorized AI agent and whether its instructions have been tampered with.
ATP was authored by OTT Cybersecurity's Lyrie.ai research team. The protocol documentation is publicly available at lyrie.ai/research, and the reference implementation has been released under the MIT license on github.com/OTT-Cybersecurity-LLC/lyrie-ai. Guy Sheetrit, CEO and Founder of OTT Cybersecurity LLC, stated: "Every AI agent on the internet today is a stranger. You don't know who it is, what it's authorized to do, or if it has been tampered with. ATP is the protocol that changes that."
On the same day, the company announced another milestone: it was accepted as one of the first members of Anthropic's Cyber Verification Program (CVP). CVP is a framework used by Anthropic to verify legitimate dual-use cybersecurity operators. This acceptance will support Lyrie in conducting vulnerability research, offensive security tool development, and red teaming workflows on Claude AI infrastructure, with related activities subject to Anthropic's applicable security and safety policies. Sheetrit commented on this: "Being one of the first companies accepted into Anthropic's Cyber Verification Program validates everything we do. Lyrie is not a security tool running alongside AI; it is the security layer upon which AI operates."
Lyrie.ai is an autonomous cybersecurity platform under OTT Cybersecurity LLC, positioned as an integrated offensive and defensive security system for the AI era. The platform's core features include: lyrie hack—a single command to execute a seven-stage autonomous penetration test, generating proof-of-concept exploit code and code-level remediation plans; adversarial AI red teaming based on H200 GPU infrastructure, supporting GCG and AutoDAN workflows; coverage of the OWASP Agentic Security Initiative (ASI) 2026 threat catalog; and an autonomous zero-day vulnerability discovery workflow achieved through Omega-Suite binary analysis.
Enterprise and government sectors are deploying autonomous AI agents at an unprecedented pace. These agents can read emails, write code, execute transactions, sign contracts, and operate within sensitive systems on behalf of human operators. However, an enterprise-grade security model for such agents has yet to emerge. ATP aims to mirror the role of the SSL/TLS protocol as a trust layer for the internet, intending to become the infrastructure-level trust standard within the autonomous AI economy.
In addition to the ATP release and CVP acceptance, Lyrie.ai simultaneously completed a $2 million Pre-Seed funding round and officially exited stealth mode. The funds will be used for security research team expansion, advancing the ATP IETF standard submission process, and deepening enterprise and government collaborations. The company is preparing for a Series A funding round to extend platform deployment to broader enterprise and government markets.
Lyrie.ai is also exploring the establishment of verification pathways similar to CVP with other leading AI labs, as part of its mission to build trusted security infrastructure for autonomous AI systems. ATP is a royalty-free protocol, which any organization can adopt free of charge without paying patent licensing fees.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
