Deconfusing AI-based IAM & IAM for AI Capabilities
Exploring the distinctions between AI-based Identity and Access Management and IAM for AI capabilities. How do these concepts intersect, and what are their implications?
Read moreAccess control has proven central to the resilience of digital infrastructure, with security policies created for sensitive documents, cloud resources, and more. We extend this paradigm by natively handling AI capabilities as tangible resources, helping organizations deploy agents and assistants with confidence.
Current methods of managing access to generative capabilities are either too permissive or too restrictive. For example, guardrails are susceptible to jailbreaks, while unlearning leaves value on the table. We highlight a blindspot of the status quo: many capabilities can be used for both good and bad.
We propose a new approach for managing access to generative capabilities, where the messy, complex issue of misuse is reduced to a familiar access problem. Building on decades of research in information security, we delegate decisions about access to a external, battle-tested access control module.
From mere assistants to autonomous agents, and from playful experiments to emerging standards, the development of generative capabilities is evolving rapidly. We develop timeless infrastructure meant to strategically accommodate such paradigm shifts, and catalyze yet others.
Stay up-to-date with resources, announcements, and opportunities at the intersection of AI, security, and infrastructure.
Exploring the distinctions between AI-based Identity and Access Management and IAM for AI capabilities. How do these concepts intersect, and what are their implications?
Read moreWe’re excited to share Pinboard, a command-line tool that streamlines workflows for developers working with generative systems. Learn how Pinboard can help you manage file references, request in-place file updates, and boost productivity in codebase-level development tasks.
Read moreChallengers are individuals who can push frontier models to their absolute limits. They're passionate about the integrity of digital, biological, and social systems, and are stress-testing our simulators across cybersecurity, biosecurity, and beyond — for fun and profit.