diff --git a/extensions/positron-python/pythonFiles/positron/positron_ipykernel/data_explorer.py b/extensions/positron-python/pythonFiles/positron/positron_ipykernel/data_explorer.py
index b920b93a76e..05eccba168b 100644
--- a/extensions/positron-python/pythonFiles/positron/positron_ipykernel/data_explorer.py
+++ b/extensions/positron-python/pythonFiles/positron/positron_ipykernel/data_explorer.py
@@ -297,10 +297,6 @@ def __init__(
         # performance.
         self._search_schema_last_result: Optional[Tuple[str, List[ColumnSchema]]] = None
 
-        # squelch a warning from pandas 2.2.0 about the use below of
-        # fillna
-        pd_.set_option("future.no_silent_downcasting", True)
-
         # Putting this here rather than in the class body before
         # Python < 3.10 has fussier rules about staticmethods
         self._SUMMARIZERS = {
@@ -560,9 +556,12 @@ def _eval_filter(self, filt: RowFilter):
                 elif params.type == SearchFilterParamsType.EndsWith:
                     mask = col.str.endswith(term)
 
+        assert mask is not None
+
         # Nulls are possible in the mask, so we just fill them if any
         if mask.dtype != bool:
-            mask = mask.fillna(False).infer_objects(copy=False)
+            mask[mask.isna()] = False
+            mask = mask.astype(bool)
 
         return mask.to_numpy()
 
diff --git a/extensions/positron-python/pythonFiles/positron/positron_ipykernel/tests/test_data_explorer.py b/extensions/positron-python/pythonFiles/positron/positron_ipykernel/tests/test_data_explorer.py
index 6da925b861d..26c04fb157b 100644
--- a/extensions/positron-python/pythonFiles/positron/positron_ipykernel/tests/test_data_explorer.py
+++ b/extensions/positron-python/pythonFiles/positron/positron_ipykernel/tests/test_data_explorer.py
@@ -832,7 +832,8 @@ def test_pandas_filter_search(dxf: DataExplorerFixture):
     ]
 
     for search_type, column_index, term, cs, mask in cases:
-        ex_table = df[mask.fillna(False)]
+        mask[mask.isna()] = False
+        ex_table = df[mask.astype(bool)]
         dxf.check_filter_case(
             df,
             [