AI as an inventor – What triggered it and where are we now?

Artificial Intelligence (AI) is present in every field nowadays from art, music, and finance to medicine. But there is still resistance to AI being officially accepted as an inventor. Let’s examine the recent developments regarding this topic.

AI-generated inventions in the last 4 years

  1. DABUS case:

    In 2018, Stephen Thaler filed two European patent applications which he claimed DABUS (Device for the Autonomous Bootstrapping of Unified Sentience) had invented. DABUS is an AI system that conceived 2 inventions for which Thaler filed patent applications in 17 jurisdictions listing it as the inventor. These were the first cases that challenged traditional patent laws of who can be recognized as an inventor. As Ryan Abbott, a University of Surrey law professor and author said, “We’re moving into a new paradigm where not only do people invent, people build artificial intelligence that can invent.”

    In 2019, USPTO rejected the patent applications saying that only humans can be recognized as inventors under the Patent Act. Between 2018 and 2019, Thaler filed patent applications in other major patent offices such as the European Patent Office (EPO), United Kingdom Intellectual Property Office (UKIPO), and IP Australia as well, but got rejected again for the same reason. However in 2021, the South African Patent Office agreed to list the AI machine DABUS as an inventor. This is putting pressure on the US and Europe to re-assess their decision on the matter.

  1. Halicin:

    In 2020, a machine learning algorithm at MIT helped researchers in developing a potent antibiotic, Halicin, which is effective against many pathogens. The computer model screened millions of chemical compounds to pick out potential antibiotics that kill bacteria using different mechanisms than those of existing drugs. Therefore, AI played a vital role in developing this drug.

    Although they haven’t filed a patent for it yet, researchers say that it is a promising way of using computational methods to discover and predict the properties of potential drugs.

The future of AI as a patent inventor

The fact that AI will be involved in numerous inventions within a few years is posing a threat to patent systems. Well-defined decisions regarding AI as an inventor are the need of the hour, especially because its’ market is projected to grow at a rate of 35-36% in the next 5 years (2022-2027).

Granting patents for AI-generated inventions will drive national innovation potential and economic growth. Additionally, it will encourage professionals who develop and use AI as it helps in creating valuable and life-saving inventions.

Currently, courts are wrestling with the problem of AI as an inventor, particularly because patent applications listing it as an inventor have been filed in more than 100 countries. AI is pushing the boundaries of existing patent laws, and many are suggesting that national governments should create customized legislation to support and protect AI-generated innovations and inventions. Policies should be designed to protect AI inventiveness and laws should make sure these policies are followed.

Whether or not AI can become a patent inventor is a developing story. Here are a few links to the latest announcements by patent offices:

  • AI cannot be named as inventor on patent applications –
  • Artificial intelligence – Inventorship of AI inventions –

PatSeer is an IP intelligence software with multiple AI-driven tools to assist users with searching, patent analytics, reporting, collaborating, and managing patent data projects. Its AI and ML capabilities collect, curate, and transform raw data into ready to use information to fulfill your IP intelligence needs. It leverages an AI-driven NLP engine – ReleSense™ to power its semantic patent search.

Start your AI-powered patent search in PatSeer