diff --git a/python/packages/agbench/src/agbench/run_cmd.py b/python/packages/agbench/src/agbench/run_cmd.py
index a8a8161c5..a8a781229 100644
--- a/python/packages/agbench/src/agbench/run_cmd.py
+++ b/python/packages/agbench/src/agbench/run_cmd.py
@@ -284,14 +284,8 @@ def get_scenario_env(token_provider: Optional[Callable[[], str]] = None, env_fil
 
     ## Support Azure auth tokens
     azure_openai_ad_token = os.environ.get("AZURE_OPENAI_AD_TOKEN")
-    if not azure_openai_ad_token and token_provider:
+    if azure_openai_ad_token is None and token_provider is not None:
         azure_openai_ad_token = token_provider()
-    if not azure_openai_ad_token:
-        azure_token_provider = get_azure_token_provider()
-        if azure_token_provider:
-            azure_openai_ad_token = azure_token_provider()
-        else:
-            logging.warning("No Azure AD token provider found. Azure AD token not set.")
     if azure_openai_ad_token is not None and len(azure_openai_ad_token.strip()) > 0:
         env["AZURE_OPENAI_AD_TOKEN"] = azure_openai_ad_token
 
@@ -888,6 +882,12 @@ def run_cli(args: Sequence[str]) -> None:
         help="The number of parallel processes to run (default: 1).",
         default=1,
     )
+    parser.add_argument(
+        "-a",
+        "--azure",
+        action="store_true",
+        help="Use Azure identity to pass an AZURE_OPENAI_AD_TOKEN to the task environment. This is necessary when using Azure-hosted OpenAI models rather than those hosted by OpenAI.",
+    )
     parser.add_argument(
         "-e",
         "--env",
@@ -972,7 +972,9 @@ def run_cli(args: Sequence[str]) -> None:
                 )
 
     # Get the Azure bearer token generator if a token wasn't provided and there's any evidence of using Azure
-    azure_token_provider = get_azure_token_provider()
+    azure_token_provider = None
+    if parsed_args.azure:
+        azure_token_provider = get_azure_token_provider()
 
     # Run the scenario
     if parsed_args.parallel > 1: