OpenClaw & Moltbook: How to Control Access of AI Agents

​NOTE: You will receive a separate calendar invite from our webinar platform upon registering.

​OpenClaw just proved it overnight: The old access model does not work to govern and control access of AI agents. 

​So, as companies roll out agents at rapid speed (and agents are on a good way to outnumber humans in terms of access), the new billion dollar question is:

​How to govern access of AI agents?

​The traditional human access world was: 

• ​Identity: One known identity (the human) 

• ​Access logic: Access to one system 

• ​Permissions: Static

• ​Auth: Usually via SSO

​With agents, this model no longer applies:

• ​Identity: Non-human (often with sub agents)

• ​Access logic: Multiple system-to-system accesses per agent

• ​Permissions: Should be dynamic/job-based (not the case today, big issue!) 

• ​Auth: Today mostly via API key or OAuth (but that can’t be the solution)

​In this session, Johannes Keienburg (CEO, Cakewalk) and Mackenzie Jackson (Developer and Security Advocate, Aikido Security) unpack what OpenClaw and Moltbook really tell us about the future of access and how to adopt AI agents without losing control.

​What You’ll Learn

• ​Why AI agents and other non-human identities are the fastest-growing access type

• ​How static permission scopes lead to OpenClaw-style incidents

• ​What it means to govern AI agent access in a dynamic (job specific!), auditable way

• ​How to get visibility into AI agents, system-to-system access, and permission scope

• ​How to stay audit-ready by default as agents scale faster than humans

​This is a practical session for IT, Security, and Ops leaders who need to enable AI agents without turning access governance into a guessing game and who want a model that scales with autonomy, not headcount.

​Even if you can't attend, sign up anyway and we'll send you the recording.