diff --git a/polars/polars-core/src/chunked_array/ops/aggregate/mod.rs b/polars/polars-core/src/chunked_array/ops/aggregate/mod.rs
index 12584c3baa5f..3db7a22fee53 100644
--- a/polars/polars-core/src/chunked_array/ops/aggregate/mod.rs
+++ b/polars/polars-core/src/chunked_array/ops/aggregate/mod.rs
@@ -223,7 +223,6 @@ where
 
 /// Booleans are casted to 1 or 0.
 impl BooleanChunked {
-    /// Returns `None` if the array is empty or only contains null values.
     pub fn sum(&self) -> Option<IdxSize> {
         Some(if self.is_empty() {
             0
diff --git a/polars/polars-core/src/chunked_array/ops/is_in.rs b/polars/polars-core/src/chunked_array/ops/is_in.rs
index 7c408f977ebf..d6979b3476db 100644
--- a/polars/polars-core/src/chunked_array/ops/is_in.rs
+++ b/polars/polars-core/src/chunked_array/ops/is_in.rs
@@ -269,7 +269,13 @@ impl IsIn for BooleanChunked {
             DataType::Boolean => {
                 let other = other.bool().unwrap();
                 let has_true = other.any();
-                let has_false = !other.all();
+                let nc = other.null_count();
+
+                let has_false = if nc == 0 {
+                    !other.all()
+                } else {
+                    !(other.sum().unwrap() as usize + nc) == other.len()
+                };
                 Ok(self.apply(|v| if v { has_true } else { has_false }))
             }
             _ => polars_bail!(opq = is_in, self.dtype(), other.dtype()),
diff --git a/py-polars/tests/unit/operations/test_is_in.py b/py-polars/tests/unit/operations/test_is_in.py
index 8d283e4d8a6f..81a01b4245c4 100644
--- a/py-polars/tests/unit/operations/test_is_in.py
+++ b/py-polars/tests/unit/operations/test_is_in.py
@@ -28,10 +28,10 @@ def test_struct_logical_is_in() -> None:
 
 
 def test_is_in_bool() -> None:
-    bool_value_to_filter_on = {True, None}
+    vals = [True, None]
     df = pl.DataFrame({"A": [True, False, None]})
-    assert df.filter(pl.col("A").is_in(bool_value_to_filter_on)).to_dict(False) == {
-        "A": [True, False]
+    assert df.select(pl.col("A").is_in(vals)).to_dict(False) == {
+        "A": [True, False, None]
     }