Kontext raises $4m for AI agent authorisation controls

Munich-based AI security startup Kontext has raised $4m in a funding round led by 42CAP, with participation from Andreessen Horowitz’s crypto startup accelerator a16z CSX and High-Tech Gründerfonds Management GmbH. The company plans to use the money to expand its engineering team.

Kontext’s software sits between an AI agent and the systems it acts on, checking each requested action against security policy before it runs. The company said its task-aware approach weighs the agent’s identity, the resource it wants to reach and the job it was given, so an agent assigned to fix a bug can read a code repository but cannot send that code to an outside service.

Teams can run the product in an observe mode before switching on enforcement, and every decision is logged. Kontext currently works with coding agents including Anthropic’s Claude Code and OpenAI’s Codex. Individual developers can use it free of charge, with paid team plans starting at $149 a month.