diff --git a/qiskit_neko/aer_plugin.py b/qiskit_neko/aer_plugin.py
index 5e86239..c1a4ce2 100644
--- a/qiskit_neko/aer_plugin.py
+++ b/qiskit_neko/aer_plugin.py
@@ -12,9 +12,6 @@
 
 """Qiskit Aer default backend plugin."""
 
-import qiskit_aer as aer
-from qiskit_ibm_runtime.fake_provider import FakeProviderForBackendV2
-
 from qiskit_neko import backend_plugin
 
 
@@ -22,6 +19,8 @@ class AerBackendPlugin(backend_plugin.BackendPlugin):
     """A backend plugin for using qiskit-aer as the backend."""
 
     def __init__(self):
+        from qiskit_ibm_runtime.fake_provider import FakeProviderForBackendV2
+
         super().__init__()
         self.mock_provider = FakeProviderForBackendV2()
         self.mock_provider_backend_names = set()
@@ -46,6 +45,8 @@ def get_backend(self, backend_selection=None):
             the defailt settings.
         :raises ValueError: If an invalid backend selection string is passed in
         """
+        import qiskit_aer as aer
+
         if backend_selection is None:
             return aer.AerSimulator()
         if backend_selection.startswith("method="):
diff --git a/qiskit_neko/tests/machine_learning/test_neural_network_classifier.py b/qiskit_neko/tests/machine_learning/test_neural_network_classifier.py
index 6fc09cf..f441963 100644
--- a/qiskit_neko/tests/machine_learning/test_neural_network_classifier.py
+++ b/qiskit_neko/tests/machine_learning/test_neural_network_classifier.py
@@ -13,14 +13,10 @@
 """Tests for quantum neural networks classifier."""
 import numpy as np
 from ddt import ddt, data
+import unittest
 
+from qiskit import __version__ as qiskit_version
 from qiskit.primitives import StatevectorSampler as ReferenceSampler
-from qiskit_aer.primitives import Sampler as AerSampler
-
-from qiskit_machine_learning.optimizers import COBYLA
-from qiskit_machine_learning.utils import algorithm_globals
-
-from qiskit_machine_learning.algorithms.classifiers import VQC
 
 from qiskit_neko import decorators
 from qiskit_neko.tests import base
@@ -30,15 +26,31 @@
 class TestNeuralNetworkClassifierOnPrimitives(base.BaseTestCase):
     """Test adapted from the qiskit_machine_learning tutorials."""
 
+    @unittest.skipIf(
+        tuple(map(int, qiskit_version.split(".")[:2])) >= (2, 0),
+        "Skipping test until Qiskit Aer and Machine Learning are ready for Qiskit 2.0. "
+        "Tracked in: https://github.com/Qiskit/qiskit-neko/issues/54",
+    )
     def setUp(self):
-        super().setUp()
+        from qiskit_aer.primitives import Sampler as AerSampler
 
+        super().setUp()
         self.samplers = dict(reference=ReferenceSampler(), aer=AerSampler(run_options={"seed": 42}))
 
+    @unittest.skipIf(
+        tuple(map(int, qiskit_version.split(".")[:2])) >= (2, 0),
+        "Skipping test until Qiskit Aer and Machine Learning are ready for Qiskit 2.0. "
+        "Tracked in: https://github.com/Qiskit/qiskit-neko/issues/54",
+    )
     @decorators.component_attr("terra", "aer", "machine_learning")
     @data("reference", "aer")
     def test_neural_network_classifier(self, implementation):
         """Test the execution of quantum neural networks using VQC."""
+
+        from qiskit_machine_learning.optimizers import COBYLA
+        from qiskit_machine_learning.utils import algorithm_globals
+        from qiskit_machine_learning.algorithms.classifiers import VQC
+
         rng = np.random.default_rng(seed=42)
         algorithm_globals.random_seed = 42
 
diff --git a/qiskit_neko/tests/machine_learning/test_neural_networks_primitives.py b/qiskit_neko/tests/machine_learning/test_neural_networks_primitives.py
index a05994a..55fe28c 100644
--- a/qiskit_neko/tests/machine_learning/test_neural_networks_primitives.py
+++ b/qiskit_neko/tests/machine_learning/test_neural_networks_primitives.py
@@ -14,16 +14,16 @@
 
 import numpy as np
 from ddt import ddt, data, unpack
+import unittest
+
+from qiskit import __version__ as qiskit_version
 from qiskit import QuantumCircuit
 from qiskit.circuit import Parameter
 from qiskit.quantum_info import SparsePauliOp
-
 from qiskit.primitives import (
     StatevectorSampler as ReferenceSampler,
     StatevectorEstimator as ReferenceEstimator,
 )
-from qiskit_aer.primitives import Sampler as AerSampler, Estimator as AerEstimator
-from qiskit_machine_learning.neural_networks import SamplerQNN, EstimatorQNN
 
 from qiskit_neko import decorators
 from qiskit_neko.tests import base
@@ -33,9 +33,16 @@
 class TestNeuralNetworksOnPrimitives(base.BaseTestCase):
     """Test adapted from the qiskit_machine_learning tutorials."""
 
+    @unittest.skipIf(
+        tuple(map(int, qiskit_version.split(".")[:2])) >= (2, 0),
+        "Skipping test until Qiskit Aer and Machine Learning are ready for Qiskit 2.0. "
+        "Tracked in: https://github.com/Qiskit/qiskit-neko/issues/54",
+    )
     def setUp(self):
         super().setUp()
 
+        from qiskit_aer.primitives import Sampler as AerSampler, Estimator as AerEstimator
+
         self.input_params = [Parameter("x")]
         self.weight_params = [Parameter("w")]
         self.circuit = QuantumCircuit(1)
@@ -48,11 +55,18 @@ def setUp(self):
             reference=ReferenceEstimator(seed=42), aer=AerEstimator(run_options={"seed": 42})
         )
 
+    @unittest.skipIf(
+        tuple(map(int, qiskit_version.split(".")[:2])) >= (2, 0),
+        "Skipping test until Qiskit Aer and Machine Learning are ready for Qiskit 2.0. "
+        "Tracked in: https://github.com/Qiskit/qiskit-neko/issues/54",
+    )
     @decorators.component_attr("terra", "aer", "machine_learning")
     @data(["reference", 2], ["aer", 1])
     @unpack
     def test_sampler_qnn(self, implementation, decimal):
         """Test the execution of quantum neural networks using SamplerQNN."""
+        from qiskit_machine_learning.neural_networks import SamplerQNN
+
         sampler = self.samplers[implementation]
 
         qnn = SamplerQNN(
@@ -70,11 +84,18 @@ def test_sampler_qnn(self, implementation, decimal):
         np.testing.assert_array_almost_equal(input_grad, [[[-0.2273], [0.2273]]], decimal)
         np.testing.assert_array_almost_equal(weight_grad, [[[-0.2273], [0.2273]]], decimal)
 
+    @unittest.skipIf(
+        tuple(map(int, qiskit_version.split(".")[:2])) >= (2, 0),
+        "Skipping test until Qiskit Aer and Machine Learning are ready for Qiskit 2.0. "
+        "Tracked in: https://github.com/Qiskit/qiskit-neko/issues/54",
+    )
     @decorators.component_attr("terra", "aer", "machine_learning")
     @data(["reference", 2], ["aer", 1])
     @unpack
     def test_estimator_qnn(self, implementation, decimal):
         """Test the execution of quantum neural networks using EstimatorQNN."""
+        from qiskit_machine_learning.neural_networks import EstimatorQNN
+
         estimator = self.estimators[implementation]
 
         qnn = EstimatorQNN(