diff --git a/ldclient/__init__.py b/ldclient/__init__.py
index b1341f95..1536f331 100644
--- a/ldclient/__init__.py
+++ b/ldclient/__init__.py
@@ -14,8 +14,6 @@
 
 __LONG_SCALE__ = float(0xFFFFFFFFFFFFFFF)
 
-__BUILTINS__ = ["key", "ip", "country", "email", "firstName", "lastName", "avatar", "name", "anonymous"]
-
 """Settings."""
 start_wait = 5
 
diff --git a/ldclient/impl/evaluator_common.py b/ldclient/impl/evaluator_common.py
index 5e7c93f0..52a03977 100644
--- a/ldclient/impl/evaluator_common.py
+++ b/ldclient/impl/evaluator_common.py
@@ -19,8 +19,6 @@
 
 __LONG_SCALE__ = float(0xFFFFFFFFFFFFFFF)
 
-__BUILTINS__ = ["key", "secondary", "ip", "country", "email", "firstName", "lastName", "avatar", "name", "anonymous"]
-
 
 # EvalResult is used internally to hold the EvaluationDetail result of an evaluation along with
 # other side effects that are not exposed to the application, such as events generated by
diff --git a/ldclient/impl/util.py b/ldclient/impl/util.py
index 54caf9de..fda0e4fd 100644
--- a/ldclient/impl/util.py
+++ b/ldclient/impl/util.py
@@ -18,13 +18,12 @@ def timedelta_millis(delta: timedelta) -> float:
     return delta / timedelta(milliseconds=1)
 
 
-log = logging.getLogger('ldclient.util')  # historical logger name
+# historical logger name; cleaning up the SDK's logger names is tracked in SDK-2696
+log = logging.getLogger('ldclient.util')
 
 
 __LONG_SCALE__ = float(0xFFFFFFFFFFFFFFF)
 
-__BUILTINS__ = ["key", "ip", "country", "email", "firstName", "lastName", "avatar", "name", "anonymous"]
-
 __BASE_TYPES__ = (str, float, int, bool)
 
 # Maximum length for SDK keys