diff --git a/keras_core/layers/__init__.py b/keras_core/layers/__init__.py
index a1bb2ada0..c4da9a3c6 100644
--- a/keras_core/layers/__init__.py
+++ b/keras_core/layers/__init__.py
@@ -3,4 +3,7 @@ from keras_core.layers.core.dense import Dense
 from keras_core.layers.core.input_layer import Input
 from keras_core.layers.core.input_layer import InputLayer
 from keras_core.layers.layer import Layer
+from keras_core.layers.regularization.activity_regularization import (
+    ActivityRegularization,
+)
 from keras_core.layers.regularization.dropout import Dropout
diff --git a/keras_core/layers/layer.py b/keras_core/layers/layer.py
index 3ed3d4c1c..5ac23acd3 100644
--- a/keras_core/layers/layer.py
+++ b/keras_core/layers/layer.py
@@ -343,7 +343,9 @@ class Layer(Operation):
             outputs = super().__call__(*args, **kwargs)
             # Record activity regularizer loss.
             if self.activity_regularizer is not None:
-                self.add_loss(self.activity_regularizer(outputs))
+                for output in nest.flatten(outputs):
+                    if backend.is_tensor(output):
+                        self.add_loss(self.activity_regularizer(output))
 
         # TODO: Set masks on outputs
         # self._set_mask_metadata(inputs, outputs, previous_mask)
diff --git a/keras_core/layers/layer_test.py b/keras_core/layers/layer_test.py
index 6f8ba1242..48162cd7c 100644
--- a/keras_core/layers/layer_test.py
+++ b/keras_core/layers/layer_test.py
@@ -149,6 +149,11 @@ class LayerTest(testing.TestCase):
         self.assertLen(layer.losses, 1)
         self.assertAllClose(layer.losses[0], 0.01)
 
+        # KerasTensors are no op
+        layer = ActivityRegularizer(activity_regularizer="l1")
+        layer(layers.Input((2, 2)))
+        self.assertLen(layer.losses, 0)
+
     def test_add_loss(self):
         class LossLayer(layers.Layer):
             def call(self, x):
diff --git a/keras_core/layers/regularization/activity_regularization.py b/keras_core/layers/regularization/activity_regularization.py
new file mode 100644
index 000000000..28bdca9ce
--- /dev/null
+++ b/keras_core/layers/regularization/activity_regularization.py
@@ -0,0 +1,40 @@
+from keras_core import regularizers
+from keras_core.api_export import keras_core_export
+from keras_core.layers.layer import Layer
+
+
+@keras_core_export("keras_core.layers.ActivityRegularization")
+class ActivityRegularization(Layer):
+    """Layer that applies an update to the cost function based input activity.
+
+    Args:
+        l1: L1 regularization factor (positive float).
+        l2: L2 regularization factor (positive float).
+
+    Input shape:
+        Arbitrary. Use the keyword argument `input_shape`
+        (tuple of integers, does not include the samples axis)
+        when using this layer as the first layer in a model.
+
+    Output shape:
+        Same shape as input.
+    """
+
+    def __init__(self, l1=0.0, l2=0.0, **kwargs):
+        super().__init__(
+            activity_regularizer=regularizers.L1L2(l1=l1, l2=l2), **kwargs
+        )
+        self.supports_masking = True
+        self.l1 = l1
+        self.l2 = l2
+
+    def call(self, inputs):
+        return inputs
+
+    def compute_output_shape(self, input_shape):
+        return input_shape
+
+    def get_config(self):
+        base_config = super().get_config()
+        config = {"l1": self.l1, "l2": self.l2}
+        return {**base_config, **config}
diff --git a/keras_core/layers/regularization/activity_regularization_test.py b/keras_core/layers/regularization/activity_regularization_test.py
new file mode 100644
index 000000000..b9e97de9b
--- /dev/null
+++ b/keras_core/layers/regularization/activity_regularization_test.py
@@ -0,0 +1,26 @@
+import numpy as np
+
+from keras_core import layers
+from keras_core.testing import test_case
+
+
+class ActivityRegularizationTest(test_case.TestCase):
+    def test_correctness(self):
+        layer = layers.ActivityRegularization(l1=0.2, l2=0.3)
+        layer(2 * np.ones((1,)))
+        self.assertLen(layer.losses, 1)
+        self.assertAllClose(layer.losses[0], 4 * 0.3 + 2 * 0.2)
+
+    def test_basics(self):
+        self.run_layer_test(
+            layers.ActivityRegularization,
+            {"l1": 0.1, "l2": 0.2},
+            input_shape=(2, 3),
+            input_dtype="float32",
+            expected_output_shape=(2, 3),
+            expected_num_trainable_weights=0,
+            expected_num_non_trainable_weights=0,
+            expected_num_seed_generators=0,
+            expected_num_losses=1,
+            supports_masking=True,
+        )