diff --git a/charset_normalizer/__init__.py b/charset_normalizer/__init__.py
index 596898e2..f14a90b1 100644
--- a/charset_normalizer/__init__.py
+++ b/charset_normalizer/__init__.py
@@ -25,9 +25,6 @@
 from charset_normalizer.version import __version__, VERSION
 from charset_normalizer.models import CharsetMatch, CharsetMatches
 
-# Load backport unicodedata2 if available
-import charset_normalizer.hook
-
 # Backward-compatible v1 imports
 from charset_normalizer.models import CharsetNormalizerMatch
 import charset_normalizer.api as CharsetDetector
diff --git a/charset_normalizer/hook.py b/charset_normalizer/hook.py
deleted file mode 100644
index ca05b213..00000000
--- a/charset_normalizer/hook.py
+++ /dev/null
@@ -1,7 +0,0 @@
-import sys
-
-try:
-    import unicodedata2  # type: ignore
-    sys.modules['unicodedata'] = unicodedata2
-except ImportError:
-    pass
diff --git a/charset_normalizer/utils.py b/charset_normalizer/utils.py
index bb433ca2..db59739f 100644
--- a/charset_normalizer/utils.py
+++ b/charset_normalizer/utils.py
@@ -1,4 +1,8 @@
-import unicodedata
+try:
+    import unicodedata2 as unicodedata
+except ImportError:
+    import unicodedata
+
 from codecs import IncrementalDecoder
 from re import findall
 from typing import Optional, Tuple, Union, List, Set