Kontext
Munich startup providing runtime security controls that evaluate and block AI agent actions before they execute.
Visit Website ↗ + Add to Compare Claim This CompanyOverview
Kontext, based in Munich and co-founded by Jens Ernstberger and Michel Osswald, launched publicly on September 24, 2026 with a $4M seed round led by 42CAP with a16z CSX and HTGF.
Its software sits between AI agents and the tools they use, evaluating each action in real time against identity, task context, policy and risk signals. Teams can start in observe mode and move to enforcement. It supports coding agents such as Claude Code and Codex; individual use is free and team plans start at $149 a month.
Innovation Matrix Assessment
Task-aware runtime authorization for agents was shipped at launch with coding-agent support.
Observe-then-enforce rollout is practical for security teams.
$4M seed from credible investors; no named customers.
Task-context decisions go beyond static permissions, though others pursue similar agent runtime controls.
No independent tests or customer evidence found.
Governing autonomous agent actions looks durable as agents spread.
Why CISOs Should Care
Gives security teams a pre-execution control over what AI agents may do with credentials and data.
What Makes It Different
Decides per action using the agent's task context rather than only static role permissions.
The Matrix Verdict
57/100 — INCREMENTAL INNOVATOR
Kontext is an Incremental Innovator: strong theme and thoughtful design, but only a seed-stage product without independent evidence.
Editorial Note: Claims vs. Verified Findings
Capabilities come from launch material. Founding year (2026) and the website domain are not confirmed in sources; coverage disagrees on Munich vs Berlin, with most sources saying Munich.
Sources
Alternatives to Kontext
Adaptive Security
AI-driven platform that simulates deepfake, voice, and multichannel social-engineering attacks to train and test organizations against next-generation phishing.
Quilr
CISO ReviewedEarly-stage agentic AI security startup building a 'Service-as-Software' platform to guard against human-related breaches and secure AI agent…
Zenity
Governance and security platform for AI agents and low-code/no-code development, securing agent identity, permissions and behavior across the…
Gray Swan AI
Carnegie Mellon-linked startup building automated red-teaming and runtime guardrails that frontier AI labs use to stress-test their models…
Tenzai
An agentic AI penetration testing startup building autonomous 'AI hackers' to find and validate exploitable vulnerabilities at a…
Reco
Reco secures the "agentic ecosystem" — mapping what AI agents can access across SaaS and enterprise apps, detecting…