diff --git a/src/PostgREST/SchemaCache.hs b/src/PostgREST/SchemaCache.hs
index 2cbf5c37f..90f69e3f0 100644
--- a/src/PostgREST/SchemaCache.hs
+++ b/src/PostgREST/SchemaCache.hs
@@ -511,7 +511,8 @@ tablesSqlQuery pgVer =
           c.relname::name AS table_name,
           a.attname::name AS column_name,
           d.description AS description,
-          pg_get_expr(ad.adbin, ad.adrelid)::text AS column_default,
+  |] <> columnDefault <>
+  [q|
           not (a.attnotnull OR t.typtype = 'd' AND t.typnotnull) AS is_nullable,
               CASE
                   WHEN t.typtype = 'd' THEN
@@ -693,6 +694,16 @@ tablesSqlQuery pgVer =
   "ORDER BY table_schema, table_name"
   where
     relIsPartition = if pgVer >= pgVersion100 then " AND not c.relispartition " else mempty
+    -- detect default values on columns that have GENERATED .. AS IDENTITY
+    columnDefault =
+      if pgVer >= pgVersion100
+        then [q|
+        CASE
+           WHEN nullif(a.attidentity, '') is null
+             THEN pg_get_expr(ad.adbin, ad.adrelid)::text
+             ELSE format('nextval(%s)', quote_literal(pg_get_serial_sequence(a.attrelid::regclass::text, a.attname::text)))
+        END AS column_default,|]
+        else "pg_get_expr(ad.adbin, ad.adrelid)::text AS column_default,"
 
 
 -- | Gets many-to-one relationships and one-to-one(O2O) relationships, which are a refinement of the many-to-one's
diff --git a/test/spec/Feature/Query/InsertSpec.hs b/test/spec/Feature/Query/InsertSpec.hs
index 598990beb..fbb933d1b 100644
--- a/test/spec/Feature/Query/InsertSpec.hs
+++ b/test/spec/Feature/Query/InsertSpec.hs
@@ -495,6 +495,16 @@ spec actualPgVersion = do
               , matchHeaders = ["Preference-Applied" <:> "missing=default"]
               }
 
+        when (actualPgVersion >= pgVersion100) $
+          it "inserts a default on a generated by default as identity column" $ do
+            request methodPost "/channels?columns=id,data,slug&select=data,slug" [("Prefer", "return=representation"), ("Prefer", "missing=default")]
+                [json| { "slug": "foo" } |]
+              `shouldRespondWith`
+                [json| [{"data":{"foo": "bar"},"slug":"foo"}] |] -- id 1 was inserted here, we don't get it for idempotence in the tests
+                { matchStatus  = 201
+                , matchHeaders = ["Preference-Applied" <:> "missing=default"]
+                }
+
     it "inserts json that has duplicate keys" $ do
       request methodPost "/tbl_w_json" [("Prefer", "return=representation")]
           [json| { "data": { "a": 1, "a": 2 }, "id": 3 } |]
diff --git a/test/spec/fixtures/privileges.sql b/test/spec/fixtures/privileges.sql
index 6e2277ec5..e5fc79e9f 100644
--- a/test/spec/fixtures/privileges.sql
+++ b/test/spec/fixtures/privileges.sql
@@ -41,6 +41,14 @@ GRANT USAGE ON SEQUENCE
     , leak_id_seq
 TO postgrest_test_anonymous;
 
+DO $do$
+BEGIN
+  IF current_setting('server_version_num')::INT >= 100000 THEN
+    GRANT USAGE ON SEQUENCE channels_id_seq TO postgrest_test_anonymous;
+  END IF;
+END
+$do$;
+
 -- Privileges for non anonymous users
 GRANT USAGE ON SCHEMA test TO postgrest_test_author;
 GRANT ALL ON TABLE authors_only TO postgrest_test_author;
diff --git a/test/spec/fixtures/schema.sql b/test/spec/fixtures/schema.sql
index 3dffb555c..3ec8aa933 100644
--- a/test/spec/fixtures/schema.sql
+++ b/test/spec/fixtures/schema.sql
@@ -3110,3 +3110,15 @@ create table test.tbl_w_json(
   id int,
   data json
 );
+
+DO $do$
+BEGIN
+  IF current_setting('server_version_num')::INT >= 100000 THEN
+    CREATE TABLE test.channels (
+      id bigint GENERATED BY DEFAULT AS IDENTITY PRIMARY KEY,
+      data jsonb DEFAULT '{"foo": "bar"}',
+      slug text
+    );
+  END IF;
+END
+$do$;