0
Hey there, CyberNewswire readers!
Exciting news from San Francisco today! Archipelo has just announced Salmon, a groundbreaking Execution Verification Infrastructure (EVI) for AI agents and autonomous systems. This innovative technology is powered by a cryptographic protocol that ensures the verifiability of execution history, setting a new standard for security and control in the AI space.
The launch of Salmon comes in the wake of a recent incident involving OpenAI and Hugging Face, highlighting the urgent need for robust safeguards in the realm of AI. As AI systems become increasingly sophisticated, it is essential to have mechanisms in place that can keep up with the speed and complexity of these agents.
AI agents are now capable of a wide range of actions, from invoking tools to modifying infrastructure, often operating across multiple actors and systems. Salmon addresses this challenge by capturing execution as signed events and recording the state transitions they produce, providing valuable evidence for investigation, detection, and response.
Backed by industry giants like Dell Technologies Capital and notable investors including Zoom CEO Eric Yuan, Salmon is poised to revolutionize the way we approach AI security, safety, and governance.
Why is Execution Verification crucial for AI Security and Control?
Identity, authorization, runtime controls, and observability are all vital components of AI governance. However, Execution Verification goes a step further by answering the question: Can we independently verify what executed? Salmon provides the answer.
By capturing execution as signed events and linking them into a Verifiable Execution Record, Salmon ensures that the lineage between execution and resulting state is preserved, creating a valuable source of machine-consumable execution evidence.
Meet Matthew Wise, the mastermind behind Salmon:
“You cannot control autonomous systems without verifiable evidence of their execution,” says Matthew Wise, Creator & Protocol Architect of Salmon and CEO of Archipelo. “AI safety is becoming an execution problem, and Salmon is here to bridge that gap. By capturing execution evidence in real-time and making it cryptographically verifiable, we empower AI systems to operate safely and responsibly.”
Salmon is not just a tool for supervision and investigation; it is a game-changer for accountability and transparency in the world of autonomous systems.
Why do Autonomous Systems need Execution Evidence?
As AI systems gain more autonomy and authority, the need for verifiable execution history becomes increasingly critical. With Salmon, AI safety, control, security, and governance systems have access to a reliable source of evidence that can strengthen their oversight and decision-making processes.
“Agent autonomy and productivity are key benefits of AI, but they also introduce new challenges,” explains Bill Tai, an investor in Archipelo. “Salmon’s execution evidence is a game-changer for ensuring that AI agents operate within the boundaries of safety, control, and governance.”
From its origins in Developer Security Posture Management to its current focus on execution history and state lineage, Salmon is a testament to Archipelo’s commitment to innovation and excellence in the AI space.
About Salmon:
Salmon is your go-to solution for verifiable execution history in autonomous multi-actor software systems. Visit salmon.systems to learn more about how Salmon is shaping the future of AI security and control.
Get in touch:
Stan Kapusta
Archipelo, Inc
[email protected]
