diff --git a/ludwig/automl/base_config.py b/ludwig/automl/base_config.py
index 1358490f27e..375d66c6480 100644
--- a/ludwig/automl/base_config.py
+++ b/ludwig/automl/base_config.py
@@ -344,6 +344,8 @@ def infer_type(
     :return: (str) feature type
     """
     num_distinct_values = field.num_distinct_values
+    if num_distinct_values == 0:
+        return CATEGORY
     distinct_values = field.distinct_values
     if num_distinct_values <= 2 and missing_value_percent == 0:
         # Check that all distinct values are conventional bools.
@@ -385,6 +387,9 @@ def should_exclude(idx: int, field: FieldInfo, dtype: str, row_count: int, targe
     if field.name in targets:
         return False
 
+    if field.num_distinct_values == 0:
+        return True
+
     distinct_value_percent = float(field.num_distinct_values) / row_count
     if distinct_value_percent == 1.0:
         upper_name = field.name.upper()
diff --git a/tests/ludwig/automl/test_base_config.py b/tests/ludwig/automl/test_base_config.py
index 11cbe73d92f..4f68885ace7 100644
--- a/tests/ludwig/automl/test_base_config.py
+++ b/tests/ludwig/automl/test_base_config.py
@@ -16,6 +16,7 @@
     (2, ['1.5', '3.7'], 0, 0, 0.1, NUMERICAL),
     (ROW_COUNT, [], 3, 0, 0.0, TEXT),
     (ROW_COUNT, [], 1, ROW_COUNT, 0.0, IMAGE),
+    (0, [], 0, 0, 0.0, CATEGORY),
 ])
 def test_infer_type(num_distinct_values, distinct_values, avg_words, img_values, missing_vals, expected):
     field = FieldInfo(
@@ -36,6 +37,7 @@ def test_infer_type(num_distinct_values, distinct_values, avg_words, img_values,
     (0, ROW_COUNT, TEXT, 'name', False),
     (0, ROW_COUNT, NUMERICAL, TARGET_NAME, False),
     (0, ROW_COUNT - 1, NUMERICAL, 'id', False),
+    (0, 0, CATEGORY, 'empty_col', True),
 ])
 def test_should_exclude(idx, num_distinct_values, dtype, name, expected):
     field = FieldInfo(