diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
new file mode 100644
index 0000000000000000000000000000000000000000..2b17bfa4da98368d2bfbeaa7eb6020458251d44e
--- /dev/null
+++ b/CONTRIBUTING.md
@@ -0,0 +1,26 @@
+<!--
+    SPDX-FileCopyrightText: Copyright 2024 Arm Limited and/or its affiliates <open-source-office@arm.com>
+
+    SPDX-License-Identifier: Apache-2.0
+-->
+
+Contributions are welcome with exceptions to certain areas.
+
+# Requirements
+
+For contributions, a Developer Certificate of Origin (DCO) is required to certify its origin and the process is managed by [DCO v1.1](https://developercertificate.org/)
+
+The agreement to DCO can be notified by the 'Signed-off-by' message in the commit message using your real name and e-mail.
+An example is as below
+
+`Signed-off-by: Name Surname <name.surname@example.com>`
+
+# Exempted from Contributions
+
+The following two folders in the main directory are exempted from contributions.
+
+1. kai
+1. test
+
+The micro-kernels in kai folder are primarily auto generated and the source of that is not planned to be made open source. For any
+changes there, please raise an issue.
diff --git a/README.md b/README.md
index c63c1c8e83fb452595ae219c0faf76f369a96acf..f480734bb8233db47e535697f94403aca94c5b1d 100644
--- a/README.md
+++ b/README.md
@@ -193,10 +193,6 @@ KleidiAI offers performance-critical operators for ML, like matrix multiplicatio
 
 <em>The micro-kernel does not use any internal threading mechanism</em>. However, the micro-kernel's API is designed to allow the computation to be carried out only on specific areas of the output tensor. Therefore, this mechanism is sufficient to split the workload on parallel threads. More information on dispatching the micro-kernels among different threads will be available soon.
 
-<h1> How to contribute </h1>
-
-KleidiAI is not currently accepting contributions during the bring up phase.
-
 <h1> License </h1>
 
 [Apache-2.0](LICENSES/Apache-2.0.txt).