Kontext raises $4M for runtime security platform for AI agents

AI security firm Kontext has secured $4 million in funding to bolster its runtime security platform catering to AI agents. The investment was spearheaded by 42CAP, with contributions from a16z CSX and HTGF. Founded by Jens Ernstberger and Michel Osswald, Kontext's software aims to enhance organizations' visibility and control over AI agents' activities.

These agents, which have evolved from chat interfaces and software development to workplace environments, can now write code, access files, and operate using company credentials.

As businesses grant agents increased access and autonomy, Kontext's platform steps in to regulate what actions agents are authorized to perform while active. The company's software acts as an intermediary between AI agents and the tools and systems they interact with. It scrutinizes agent activity in real-time against security policies and risk signals, factoring in the agent's identity, requested action, target resource, and assigned task.

Despite proper authentication of agents and approval of their tools, they can still execute unauthorized actions. Kontext addresses this by linking identity with task context and policy to determine an agent's authorization before the action occurs. Jens Ernstberger, co-founder of Kontext, explained that teams can first deploy the platform in an observe mode to comprehend agent behavior, identify potentially risky activity, and understand how security policies would apply without disrupting their workflow.

When enforcement is activated, the platform can halt unauthorized actions before they're executed and maintain a documented record of each decision.

The funding will be allocated towards expanding Kontext's engineering team, further refining its runtime enforcement platform, and assisting customers deploying AI agents with greater control and visibility.