ALEXANDRIA, Va., Sept. 15 -- United States Patent no. 12,737,543, issued on Sept. 15, was assigned to Morgan Stanley Services Group Inc. (New York).

"Watermarking large language model outputs" was invented by Timothy S. Eng (Armonk, N.Y.), Pradeep Sundaram (Princeton Junction, N.J.) and Christine J. Tu (Cranbury, N.J.).

According to the abstract* released by the U.S. Patent & Trademark Office: "Systems and methods for implementing a post-hoc, low barrier to entry and low-cost method and system of watermarking large language model (LLM) outputs in an enterprise environment are provided. An embodiment of the present invention is directed to a method and system for embedding a unique invisible watermark into a text output of a large language...