KONTEXT Security has launched publicly with $4 million in funding from a round led by 42CAP, with additional backing from a16z CSX and HTGF. The Munich-based startup was co-founded by Jens Ernstberger and Michel Osswald and is developing a runtime security platform designed to give organisations visibility and control over AI agents’ actions.
The platform is deployed between AI agents and the systems and tools they access. It evaluates requests in real time using the agent’s identity, assigned task, target resource and proposed action, then applies security policies to determine whether the action should proceed. Organisations can initially use it in observation mode to understand agent behaviour, identify risks and review potential policy outcomes before enabling enforcement to block unauthorised actions. The system also records agent activity and the decisions it makes, together with relevant context.
Chief executive Ernstberger said authentication and approved tools alone do not guarantee that an AI agent will take an authorised action, arguing that organisations need a control point immediately before an action occurs. Kontext said it plans to use the funding to expand its engineering team and continue developing the runtime enforcement platform. The report does not identify any specific incident, vulnerability or confirmed exploitation linked to the company’s technology.