Artificial Intelligence: Intellectual Property Policy Considerations

Coresearch will exhibit new products that utilize AI, including products used in its name creation, clearance process, and online band protection services.

Entrupy will be demonstrating its suite of AI technology products, which includes a handheld device and Entrupy–configured iPod that enables the detection of counterfeit goods. The company has generated “fingerprints” of over 45 million authentic and fake items. Based on the captured microscopic images, Entrupy’s software uses machine-learning algorithms to predict whether an imaged object is genuine or counterfeit.

Google will display several internal models used to improve efficiency, including how a machine-learning model can calculate claim breadth, how similarity technology can assist with landscaping and standards, and how cross-licensing can be improved with AI. Representatives from Google Patents will present how machine-learning models have helped to classify non-patent literature documents with Cooperative Patent Classification (CPC) codes and how similarity technology can be used to retrieve and rank search results.

Dr. Paul Morris will demonstrate the NSF’s enterprise search tool, which clusters unstructured text from documents into thematic areas. The tool’s visualizations enable the NSF to manage huge volumes of research submissions and place these documents into panels and related clusters. Additionally, the tool automatically assigns domain experts and reviewers to each submitted document. There are numerous similarities between this NSF process and the steps taken when a patent application is submitted to the USPTO.

TELES PRI will demonstrate a patent-oriented prototype of IES focusing on substantive U.S. patent law and following frameworks contained in recent decisions of the U.S. Supreme Court.

The IES prototype is a deterministic AI system that enables the drafting — semi-automatically and interactively — of robust emerging technology claimed inventions and testing them in real time. Implementation, using relevant decisions of the Court of Appeals for the Federal Circuit and/or the Patents Trial and Appeals Board, will show IES’s actual workings, for example, incorporating criteria laid forth in such decisions as DDR Holdings (regarding computer-implemented patent claims), Myriad Genetics (regarding human genes), and UC Berkeley (regarding CRISPR patents).

Modules include:

Automated Novelty Checker

Automated Evidence of Use Generator

Automated Technology Evaluator

Automated Idea Generator (Beta)

The SBIR/STTR Programs Office will exhibit materials highlighting the DoD’s SBIR and STTR programs available to innovators working in the area of machine learning and AI.

The Small Business Innovation Research (SBIR) program is a U.S. government program, coordinated by the U.S. Small Business Administration, in which federal agencies reserve a portion of their research budgets for contracts or grants to small businesses. The Small Business Technology Transfer (STTR) program funds cooperative research and development projects with small businesses in partnership with nonprofit research institutions (such as universities) to move research to the marketplace. Annually, the DoD STTR program funds more than $100 million in research. The DoD SBIR program provides more than $1 billion in research funds to small businesses each year.