To date, patent tools have been developed primarily to help patent attorneys make better decisions. Unfortunately for tool developers, the patent attorneys who are already using tools and embracing data are already making well-supported decisions. True, tools can continue to improve these decisions incrementally over the next 10 years, both in accuracy and efficiency. That said, there is now a much greater need for tools that replace attorney decisions and actions, both to eliminate the decisions and actions that are not justified by the underlying facts and also to allow data-driven attorneys to make more decisions more efficiently. The need for non-data-driven attorneys is diminishing.

As humans, we naturally think that we are making the best decisions when we look at all of the relevant facts. This poses two problems in the patent industry today:  (1) patent attorneys are not acknowledging that certain available facts are relevant in one context (i.e., litigation, with a small sample size; or law firm A’s prosecution outlook) when they are clearly relevant in another context (i.e., prosecution, with a large sample size; or all law firm prosecution outlooks); and (2) patent attorneys that do acknowledge available relevant facts now have too many facts to consider in order to make quick and efficient decisions that serve the client’s interest without compounding legal fees. What is needed in the legal industry are tools that actually make the decisions and perform the actions, allowing humans to monitor and intervene only if necessary after reviewing and understanding a machine-documented justification. These tools will allow the data-driven patent attorneys to handle even more workload and serve clients’ needs even more efficiently.

Current patent tools fall short, but we are starting to see tools such as Rowan Patents and Specifio, as well as tools native to law firms and companies, that are chipping the edges off of patent attorney tasks. These tools are still missing the machine-documented justification, and that machine-documented justification is an aspect that Oracle is beginning to build into its own patent review tools. The justification will prevent humans from incorrectly overriding the machine’s decisions. For the first time in the history of patent drafting, patent review, and patent prosecution, a machine might actually recommend a better approach than a human patent attorney, and the human patent attorney might not immediately appreciate why. Although it is not yet time to entrust the entire drafting process to a machine, there are certain decisions even within the drafting process, such as word choice and phraseology, that a machine can make more accurately and efficiently than even the best of humans.