You can add a cluster level Spark configuration to set the default catalog at startup avoiding per session configurations.

spark.databricks.sql.initial.catalog.name spark_catalog

You can use spark.sql.path.enabled = true and explicitly set PATH to include system builtin

spark.conf.set("spark.sql.path.enabled", "true")
spark.sql("SET PATH = system.builtin, system.session, spark_catalog.default")