What was once the subject of futuristic thrillers and science fiction is now an integral part of life. Artificial intelligence is currently being used in healthcare, education, and finance, and the applications of AI are becoming more impressive every day. Exponential growth and development is an inherent characteristic of the technology, and as such, it is no surprise that the emergence of this machine learning has brought about a spike in the number of patent applications with AI being the core of the subject matter.

Smooth Sailing for AI

If you were to file a patent application relating to AI, the odds are good that your patent application will wind up in either Art Unit 2121 or in Art Unit 2129. Combined, the two art units have been inundated with 16,658 patent applications, and over 70 percent of those applications have been filed in the 12 years. Patent data analysis shows that patent application filings really did increase exponentially up until their peak in 2011, and they have been waning ever so slightly since.

Of all the AI patent applications filed, over three quarters of them have issued – and it appears that the path to allowance is relatively smooth. The hardest part about pursuing an AI patent appears to be having the patience to wait an average of three years, three months, and nine days after filing your application for it to develop into a granted patent. This long wait is rather uneventful in most cases. The average AI patent application faces only 1.7 office actions throughout the course of prosecution – the first office action usually not occurring until after two years have passed.

Only a third of all AI applications face a final office action, but even those applicants who face final rejections have done well finding ways to have their patents issue. Roughly a quarter of AI patent applicants file at least one RCE, and simply filing the RCE results in allowance 39.5 percent of the time. Only a handful of AI applications (5.5 percent) have entered into the appeal process. Less than 18% of those applications ever make it to a PTAB decision, but nearly 75 percent of application that enter an appeal end up winning.

Shortcuts to Allowance

As if AI patent prosecution statistics were not already favorable enough to applicants, many patent practitioners have used prosecution strategies that yield even greater efficiency. At least one interview with a patent examiner was held in 30.6 percent of the AI applications filed so far.  Patent analysis tools show us that, in these applications, almost 60 percent of them were granted by the patent examiner immediately after the interview. However, it is arguably the 210 AI patent applicants who bought into the USPTO’s Track One prioritized examination that have the most impressive patent statistics of all.  Not only did prioritized examination reduce prosecution pendency to just over a year, but the applications filed in Track One faced an average of 0.8 fewer office actions and realized a 91.7 allowance rate. Saving time and the cost of an office action, Track One appears to be paying for itself in AI.

As a patent attorney, it is important to know the landscape of an industry before stepping into patent prosecution. Patent analysis tools provide insights that help you assess your art unit and develop the best strategy for prosecuting your patent application. LexisNexis PatentAdvisor® patent analysis software gives you access to actionable patent statistics so you can make smarter decisions. Patent prosecution with PatentAdvisor™ is a true sign of intelligence.

Better Strategic Decisions PatentAdvisor CTA