# Security policy automation

Linkerd provides a rich and expressive language for expressing a type of
security policy called _authorization policy_, which gives you control over
which components of your application are able to communicate with each other,
and under what conditions. Based on Linkerd's powerful workload identity and
authenticated via mutual TLS, authorization policies can provide a powerful
strategy for implementing _zero-trust security_.

(For a technical introduction to mutual TLS, we recommend
[A Kubernetes engineer's guide to mTLS](https://buoyant.io/mtls-guide). For a
high-level overview of zero trust, we recommend
[Zero trust network security in Kubernetes with the service mesh](https://buoyant.io/resources/zero-trust-in-kubernetes-with-linkerd).)

However, getting started with authorization policies can be difficult. BEL's
authorization policy automation feature eliminates the toil of bootstrapping by
generated policies based on **observed traffic** within a functioning
application. This allows you to quickly get functioning policies, and to keep
these policies up-to-date as your application architecture evolves.

For a walkthrough of how to try authorization policies, see our walkthrough,
[Guide: Generating authorization policies on an existing application](/buoyant-enterprise-linkerd/2.15/guides/generating-policy/).

## Learning more

- [Guide: Generating authorization policies for an existing application](/buoyant-enterprise-linkerd/2.15/guides/generating-policy/)
- [Zero trust network security in Kubernetes with the service mesh](https://buoyant.io/resources/zero-trust-in-kubernetes-with-linkerd)
- [A Kubernetes engineer's guide to mTLS](https://buoyant.io/mtls-guide)
