diff --git a/internal/fingerprint/sources_checksum.go b/internal/fingerprint/sources_checksum.go
index f1108e111c..f4aee43b0c 100644
--- a/internal/fingerprint/sources_checksum.go
+++ b/internal/fingerprint/sources_checksum.go
@@ -119,7 +119,7 @@ func (checker *ChecksumChecker) checksumFilePath(t *ast.Task) string {
 	return filepath.Join(checker.tempDir, "checksum", normalizeFilename(t.Name()))
 }
 
-var checksumFilenameRegexp = regexp.MustCompile("[^A-z0-9]")
+var checksumFilenameRegexp = regexp.MustCompile("[^A-Za-z0-9]")
 
 // replaces invalid characters on filenames with "-"
 func normalizeFilename(f string) string {
diff --git a/internal/fingerprint/sources_checksum_test.go b/internal/fingerprint/sources_checksum_test.go
index a2b35cd316..9afc25b6d8 100644
--- a/internal/fingerprint/sources_checksum_test.go
+++ b/internal/fingerprint/sources_checksum_test.go
@@ -16,6 +16,12 @@ func TestNormalizeFilename(t *testing.T) {
 		{"foo/bar/baz", "foo-bar-baz"},
 		{"foo@bar/baz", "foo-bar-baz"},
 		{"foo1bar2baz3", "foo1bar2baz3"},
+		// Characters in the ASCII range between 'Z' (90) and 'a' (97) must
+		// also be normalized. A "[A-z]" class would wrongly include these.
+		{"foo\\bar", "foo-bar"},
+		{"foo_bar", "foo-bar"},
+		{"foo[bar]baz", "foo-bar-baz"},
+		{"foo^bar`baz", "foo-bar-baz"},
 	}
 	for _, test := range tests {
 		assert.Equal(t, test.Out, normalizeFilename(test.In))