diff --git a/alembic/versions/20241127_c3458e1ef9aa_remove_unsafe_characters_summary.py b/alembic/versions/20241127_c3458e1ef9aa_remove_unsafe_characters_summary.py
new file mode 100644
index 000000000..05101d74f
--- /dev/null
+++ b/alembic/versions/20241127_c3458e1ef9aa_remove_unsafe_characters_summary.py
@@ -0,0 +1,32 @@
+"""Remove unsafe characters summary
+
+Revision ID: c3458e1ef9aa
+Revises: 272da5f400de
+Create Date: 2024-11-27 20:32:41.431147+00:00
+
+"""
+
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision = "c3458e1ef9aa"
+down_revision = "272da5f400de"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # Remove any characters that are not XML safe from the summary_text field. The code has been
+    # updated to filter out these characters, but this cleans up any existing data.
+    # https://www.postgresql.org/docs/current/functions-matching.html#FUNCTIONS-POSIX-REGEXP
+    op.execute(
+        "UPDATE works SET summary_text = regexp_replace("
+        "  summary_text, '[^\u0020-\uD7FF\u0009\u000A\u000D\uE000-\uFFFD\U00010000-\U0010FFFF]+', '', 'g'"
+        ") WHERE "
+        "summary_text ~ '[^\u0020-\uD7FF\u0009\u000A\u000D\uE000-\uFFFD\U00010000-\U0010FFFF]'"
+    )
+
+
+def downgrade() -> None:
+    # No need to do anything on downgrade.
+    pass
diff --git a/src/palace/manager/sqlalchemy/model/work.py b/src/palace/manager/sqlalchemy/model/work.py
index 9c0cc4263..b3e0fdbc9 100644
--- a/src/palace/manager/sqlalchemy/model/work.py
+++ b/src/palace/manager/sqlalchemy/model/work.py
@@ -2,11 +2,13 @@
 
 from __future__ import annotations
 
+import re
 import sys
 from collections import Counter
 from collections.abc import Sequence
 from datetime import date, datetime
 from decimal import Decimal
+from functools import cache
 from typing import TYPE_CHECKING, Any, cast
 
 import opensearchpy
@@ -609,11 +611,21 @@ def merge_into(self, other_work):
 
         other_work.calculate_presentation()
 
-    def set_summary(self, resource):
+    @classmethod
+    @cache
+    def _xml_text_sanitization_regex(cls) -> re.Pattern[str]:
+        # Source: https://stackoverflow.com/questions/8733233/filtering-out-certain-bytes-in-python
+        return re.compile(
+            r"[^\u0020-\uD7FF\u0009\u000A\u000D\uE000-\uFFFD\U00010000-\U0010FFFF]+"
+        )
+
+    def set_summary(self, resource: Resource) -> None:
         self.summary = resource
-        # TODO: clean up the content
         if resource and resource.representation:
-            self.summary_text = resource.representation.unicode_content
+            # Make sure that the summary text only contains characters that are XML compatible.
+            self.summary_text = self._xml_text_sanitization_regex().sub(
+                "", resource.representation.unicode_content
+            )
         else:
             self.summary_text = ""
         WorkCoverageRecord.add_for(self, operation=WorkCoverageRecord.SUMMARY_OPERATION)
diff --git a/tests/manager/sqlalchemy/model/test_edition.py b/tests/manager/sqlalchemy/model/test_edition.py
index a2d7cd4ec..0674b60a5 100644
--- a/tests/manager/sqlalchemy/model/test_edition.py
+++ b/tests/manager/sqlalchemy/model/test_edition.py
@@ -403,20 +403,32 @@ def test_set_summary(self, db: DatabaseTransactionFixture):
         work = db.work(presentation_edition=e)
         overdrive = DataSource.lookup(db.session, DataSource.OVERDRIVE)
 
-        # Set the work's summmary.
+        # Set the work's summary.
         l1, new = pool.add_link(
             Hyperlink.DESCRIPTION, None, overdrive, "text/plain", "F"
         )
         work.set_summary(l1.resource)
 
-        assert l1.resource == work.summary
-        assert "F" == work.summary_text
+        assert work.summary == l1.resource
+        assert work.summary_text == "F"
+
+        # Set the work's summary to a string that contains characters that cannot be
+        # represented in XML.
+        l2, new = pool.add_link(
+            Hyperlink.DESCRIPTION,
+            None,
+            overdrive,
+            "text/plain",
+            "\u0000💣ü🔥\u0001\u000C",
+        )
+        work.set_summary(l2.resource)
+        assert work.summary_text == "💣ü🔥"
 
         # Remove the summary.
         work.set_summary(None)
 
-        assert None == work.summary
-        assert "" == work.summary_text
+        assert work.summary is None
+        assert work.summary_text == ""
 
     def test_calculate_evaluate_summary_quality_with_privileged_data_sources(
         self, db: DatabaseTransactionFixture