Shaping the Future of Agent Runtime Governance
As AI agents move from experimentation into production systems, organizations are beginning to face a new class of security and governance challenges. We are forming a small cohort of founding partners to collaboratively define how this emerging layer should work.
Traditional Identity and Access Management systems were designed for human users and service identities. They were not designed to govern autonomous agents acting dynamically across APIs, SaaS systems, and other agents at runtime.
Watchlight AI is developing a new architectural layer called Agent Runtime Governance (ARG): infrastructure that provides runtime visibility, authorization, and control for AI agents operating in enterprise environments.
Program Mission
The Founding Design Partner Program brings together a select group of organizations building or deploying AI agents to collaboratively define practical approaches to governing autonomous systems at runtime.
The goal is to validate architectures, identify operational challenges, and establish governance patterns that can safely support the next generation of AI-driven systems.
Areas of Collaboration
Design partners will help explore and validate key architectural patterns, including:
These areas correspond to core components currently being developed within the Watchlight platform.
What Partners Contribute
- Feedback on architectural models and governance patterns
- Input on real-world AI agent deployment scenarios
- Evaluation of early capabilities in controlled environments
- Participation in periodic design and architecture discussions
What Partners Receive
- Early access to emerging Watchlight AI capabilities
- Direct influence on product architecture and roadmap
- Collaboration on reference architectures for agent governance
- Recognition as a Founding Design Partner
- Preferential commercial terms at general availability
Program Structure
The program is intentionally small and collaborative.
3–5
partner organizations
3–6
month collaboration period
Periodic
architecture working sessions
Shared
exploration of governance patterns
Ideal Participants
The program is intended for organizations actively exploring or deploying:
Participants may include security architecture teams, identity teams, AI platform teams, and forward-looking engineering organizations.
Request Early Access
Registering takes about two minutes. We will walk you through three steps: your details, your AI agent environment, and the program terms.
Together, we can begin defining the infrastructure needed to safely govern the next generation of AI-driven systems.