Patents on Artificial Intelligence Race between the USA and the Far East

American patent applicants dominate the German innovation landscape when it comes to artificial intelligence: 36.2 percent of the patent applications published in 2018 with effect for Germany came from the USA. The main drivers were automated driving, medical technology and robotics.

Only half as many (18.1 percent) came from Germany. Japan (13.3 percent), China (5.2 percent), the Republic of Korea (3.7 percent), France (3.3 percent) and the Netherlands (2.9 percent) followed in the country ranking. This is the result of an analysis published by the German Patent and Trade Mark Office (DPMA).

According to the data, Germany has continued to lose ground compared to the United States in recent years. In 2013, 35.1 per cent of the published patent applications came from the USA and 19.5 per cent from Germany. China, which almost tripled the number of its applications in this period and increased its share of the total number of applications from 2.1 per cent to 5.2 per cent. According to a recently published study by the Mercator Institute for China Studies, a total of 30,000 patents had been applied for from China in 2018. China is on its way to becoming a leading nation in digital key technologies.

"Even if German manufacturers still seem to be doing relatively well on their home market, the dominance of American companies is already clearly visible today. For the future, there is a race between the USA and the Far East", said DPMA President Cornelia Rudloff-Schäffer at the DPMA User Forum in Munich. "The innovative capacity of German companies and our country as a whole depends crucially on developments in the field of machine learning and artificial intelligence. We have to be careful that the gap does not widen even further on this central topic for the future".

In total, the number of patent applications effective for Germany in these key technologies has increased by 80 per cent over the past ten years - from 3 142 in 2009 to 5 654 last year.

Among the main drivers are applications for automated driving. In traffic and automotive engineering, the number of applications tripled over the same period: in 2009, the two patent authorities registered 327 applications, compared with 966 last year. The development in medical technology was similarly dynamic (+131 percent).  In robotics and general control engineering, for example, the growth rates are even higher, although the number of applications is significantly lower in absolute terms. Most registrations were in 2018 for the computer technology required for artificial intelligence.

The DPMA notes that there is no generally applicable definition for the assignment of patent documents to artificial intelligence. It is therefore not possible to draw a completely clear distinction from other fields of technology. The evaluation therefore represents a meaningful approximation.