diff --git a/CHANGELOG.md b/CHANGELOG.md
index 34a0610bd18..13c5b4e27dc 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -38,6 +38,7 @@ Also, that release drops support for Python 3.9, making Python 3.10 the minimum
 * Added support for the `out` keyword to accept a tuple, bringing ufunc signatures into alignment with those in NumPy [#2664](https://github.com/IntelPython/dpnp/pull/2664)
 * Unified public API definitions in `dpnp.linalg` and `dpnp.scipy` submodules [#2663](https://github.com/IntelPython/dpnp/pull/2663)
 * Aligned the signature of `dpnp.reshape` function with Python array API by making `shape` a required argument [#2673](https://github.com/IntelPython/dpnp/pull/2673)
+* Disallowed conversion of `dpnp.ndarray` with more than one dimension to Python scalars (`int`, `float`, `complex`) to align with NumPy 2.4.0 [#2694](https://github.com/IntelPython/dpnp/pull/2694)
 
 ### Deprecated
 
diff --git a/dpnp/dpnp_array.py b/dpnp/dpnp_array.py
index 656f099a0c4..273503c22db 100644
--- a/dpnp/dpnp_array.py
+++ b/dpnp/dpnp_array.py
@@ -194,6 +194,7 @@ def __bytes__(self):
 
     def __complex__(self, /):
         """Convert a zero-dimensional array to a Python complex object."""
+        self._check_scalar_convertible()
         return self._array_obj.__complex__()
 
     def __contains__(self, value, /):
@@ -300,6 +301,7 @@ def __eq__(self, other, /):
 
     def __float__(self, /):
         """Convert a zero-dimensional array to a Python float object."""
+        self._check_scalar_convertible()
         return self._array_obj.__float__()
 
     def __floordiv__(self, other, /):
@@ -391,6 +393,7 @@ def __index__(self, /):
 
     def __int__(self, /):
         """Convert a zero-dimensional array to a Python int object."""
+        self._check_scalar_convertible()
         return self._array_obj.__int__()
 
     def __invert__(self, /):
@@ -608,6 +611,14 @@ def __xor__(self, other, /):
         r"""Return :math:`\text{self ^ value}`."""
         return dpnp.bitwise_xor(self, other)
 
+    def _check_scalar_convertible(self):
+        """Raise if array cannot be converted to a Python scalar."""
+        if self.ndim != 0:
+            raise TypeError(
+                "Only 0-dimensional dpnp.ndarray can be converted "
+                "to a Python scalar"
+            )
+
     @staticmethod
     def _create_from_usm_ndarray(usm_ary: dpt.usm_ndarray):
         """
diff --git a/dpnp/tests/test_ndarray.py b/dpnp/tests/test_ndarray.py
index c7e9dc65b99..3a03a8079f9 100644
--- a/dpnp/tests/test_ndarray.py
+++ b/dpnp/tests/test_ndarray.py
@@ -5,6 +5,7 @@
     assert_allclose,
     assert_array_equal,
     assert_equal,
+    assert_raises,
     assert_raises_regex,
 )
 
@@ -17,6 +18,7 @@
     get_complex_dtypes,
     get_float_dtypes,
     has_support_aspect64,
+    numpy_version,
 )
 from .third_party.cupy import testing
 
@@ -530,34 +532,50 @@ def test_print_dpnp_zero_shape():
     assert result == expected
 
 
-# Numpy will raise an error when converting a.ndim > 0 to a scalar
-# TODO: Discuss dpnp behavior according to these future changes
-@pytest.mark.filterwarnings("ignore::DeprecationWarning")
-@pytest.mark.parametrize("func", [bool, float, int, complex])
-@pytest.mark.parametrize("shape", [tuple(), (1,), (1, 1), (1, 1, 1)])
-@pytest.mark.parametrize(
-    "dtype", get_all_dtypes(no_float16=False, no_complex=True)
-)
-def test_scalar_type_casting(func, shape, dtype):
-    a = numpy.full(shape, 5, dtype=dtype)
-    ia = dpnp.full(shape, 5, dtype=dtype)
-    assert func(a) == func(ia)
+class TestPythonScalarConversion:
+    @pytest.mark.parametrize("shape", [tuple(), (1,), (1, 1), (1, 1, 1)])
+    @pytest.mark.parametrize(
+        "dtype", get_all_dtypes(no_float16=False, no_complex=True)
+    )
+    def test_bool_conversion(shape, dtype):
+        a = numpy.full(shape, 5, dtype=dtype)
+        ia = dpnp.full(shape, 5, dtype=dtype)
+        assert bool(a) == bool(ia)
 
+    @pytest.mark.parametrize("shape", [tuple(), (1,), (1, 1), (1, 1, 1)])
+    @pytest.mark.parametrize(
+        "dtype", get_all_dtypes(no_float16=False, no_complex=True)
+    )
+    def test_bool_method_conversion(shape, dtype):
+        a = numpy.full(shape, 5, dtype=dtype)
+        ia = dpnp.full(shape, 5, dtype=dtype)
+        assert a.__bool__() == ia.__bool__()
 
-# Numpy will raise an error when converting a.ndim > 0 to a scalar
-# TODO: Discuss dpnp behavior according to these future changes
-@pytest.mark.filterwarnings("ignore::DeprecationWarning")
-@pytest.mark.parametrize(
-    "method", ["__bool__", "__float__", "__int__", "__complex__"]
-)
-@pytest.mark.parametrize("shape", [tuple(), (1,), (1, 1), (1, 1, 1)])
-@pytest.mark.parametrize(
-    "dtype", get_all_dtypes(no_float16=False, no_complex=True)
-)
-def test_scalar_type_casting_by_method(method, shape, dtype):
-    a = numpy.full(shape, 4.7, dtype=dtype)
-    ia = dpnp.full(shape, 4.7, dtype=dtype)
-    assert_allclose(getattr(a, method)(), getattr(ia, method)(), rtol=1e-06)
+    @pytest.mark.parametrize("func", [float, int, complex])
+    @pytest.mark.parametrize("shape", [tuple(), (1,), (1, 1), (1, 1, 1)])
+    @pytest.mark.parametrize(
+        "dtype", get_all_dtypes(no_float16=False, no_complex=True)
+    )
+    def test_non_bool_conversion(func, shape, dtype):
+        a = numpy.full(shape, 5, dtype=dtype)
+        ia = dpnp.full(shape, 5, dtype=dtype)
+        assert_raises(TypeError, func(ia))
+
+        if numpy_version() >= "2.4.0":
+            assert_raises(TypeError, func(a))
+
+    @pytest.mark.parametrize("method", ["__float__", "__int__", "__complex__"])
+    @pytest.mark.parametrize("shape", [tuple(), (1,), (1, 1), (1, 1, 1)])
+    @pytest.mark.parametrize(
+        "dtype", get_all_dtypes(no_float16=False, no_complex=True)
+    )
+    def test_non_bool_method_conversion(method, shape, dtype):
+        a = numpy.full(shape, 5, dtype=dtype)
+        ia = dpnp.full(shape, 5, dtype=dtype)
+        assert_raises(TypeError, getattr(ia, method)())
+
+        if numpy_version() >= "2.4.0":
+            assert_raises(TypeError, getattr(a, method)())
 
 
 @pytest.mark.parametrize("shape", [(1,), (1, 1), (1, 1, 1)])