Boult Wade Tennant
Bulletins » DABUS applications refused by the EPO

As was widely reported last year, a series of patent applications has been filed in the name of Dr Stephen Thaler, for inventions allegedly invented by an Artificial Intelligence (AI) called “DABUS”. These applications were filed at the US Patent and Trademark Office, the United Kingdom Intellectual Property Office (UKIPO) and the European Patent Office (EPO). The applications sparked much discussion and speculation about a central question – is it permitted to name an AI as an inventor on a patent application?

A decision was issued in early December by the UKIPO  refusing the two UK DABUS patent applications. The EPO has now decided likewise: at oral proceedings held on 25 November 2019, the EPO refused the two European DABUS applications (EP18275163.6 and EP18275174.3). The EPO has now published its written decision.

The EPO formally refused each application because the designation of DABUS as the inventor did not fulfil the necessary ‘formal requirements’ for a European patent application. Specifically, the European Patent Convention (EPC) requires that the inventor(s) for a European patent application be identified. The EPO concluded that, under the EPC, the term “inventor” refers only to a natural person. This conclusion follows consideration of the legal framework of the EPC as well as the legislative history of the EPC – in the decision, the EPO argues that there is a “clear legislative understanding” that an inventor is a natural person. Therefore, naming DABUS as the inventor was deficient.

The EPO’s reasoning extended beyond a mere statement of legislative intent – central to the reasoning of the decision was that, under the EPC, an inventor has clear legal rights, including the right to be named as such on a patent application and the right to transfer property. Such legal rights can only vest in entities with legal personality – an AI, being a non-person, is unable to exercise those rights and thus cannot be understood to be an inventor within the legal meaning of the term.

Additionally, where the applicant is not the inventor, the EPC requires that the designation of the inventor also indicates the applicant’s origin of the right to the European patent (e.g. by virtue of employment or assignment). For these two applications, the designation of inventor initially indicated that Dr Thaler acquired the right to the European patent as DABUS’s employer, but this was subsequently changed to indicate that he acquired the right by succession in title. However, the EPO concluded that an AI cannot be employed (rather, it is owned), and nor can it transfer rights to a successor in title. Therefore, again, the designation of inventor was deficient.

Representatives for Dr Thaler also advanced the argument that preventing an AI from being named as inventor meant that the presence of a procedural requirement had the effect of introducing a substantive exclusion from patentability for inventions “made” by AI systems. The EPO dismissed these arguments, reaffirming that the designation of inventor is a formal requirement which is independent from (and has no bearing on) substantive patentability requirements such as novelty and inventive step.

In issuing its decision, the EPO has not considered the issue of whether the AI has actually invented anything, or indeed who the inventor should be for the DABUS applications. Whether or not an AI can actually “invent” is, perhaps, a more philosophical/societal question which might be answered elsewhere. The decision is open to appeal, and it is possible that Dr Thaler might continue to make his case before the EPO. Regardless, given the increasing role being played by AI in innovation, it is perhaps likely we have not seen the end of the discussion on this issue. At least for now, though, the position at the EPO is clear: applicants cannot name an AI as an inventor on their European patent applications.

Relevant sectors
Aerospace
  • Commercial aviation
  • Defence and security
  • Space and satellites
  • Unmanned aerial vehicles
Artificial Intelligence and Machine Learning
Automotive
  • Autonomous vehicles
  • Electric vehicles
  • Engines
Biotechnology
  • Antibody engineering
  • Antibody manufacture and formulation
  • Bioinformatics
  • Biosimilars
  • Drug delivery
  • Gene editing (e.g. CRISPR)
  • Genomic and molecular tools and methods
  • GM crops
  • Immuno-oncology (e.g. checkpoint inhibitors; modified T cells)
  • Next generation sequencing
  • Nucleic acid synthesis
  • Personalised medicine/disease biomarkers
  • Recombinant protein production and purification
  • Stem cell therapies
  • Supplementary protection certificates (SPCs)
  • Synthetic biology
  • Therapeutic antibodies
  • Vaccinology (e.g. viral vectors; mRNA vaccines)
  • Women’s health products
Chemicals
  • Agrochemicals
  • Catalysts
  • Dispersions and colloids
  • Lubricants
  • Polymers and plastics
Communications and Networks
  • Cloud computing
  • Internet of things (IOT)
  • Oceanography, marine
  • Wired and wireless networks
Computing and Software
  • Artificial intelligence and machine learning
  • Bioinformatics
  • Blockchain and distributed ledgers
  • Communications and networks
  • Computer games
  • Data and software security, cryptography and digital rights management (DRM)
  • Data management and storage, databases and data compression
  • Digital assistants, virtual assistants and software agents
  • Fintech and adtech
  • Machine vision
  • Metaverse, virtual reality (VR) and augmented reality (AR)
  • Motor capture
  • Multimedia, audio/video processing and animation
  • Natural language processing
  • Quantum computing
  • Robotic process automation
  • Search engines
  • Signal processing
  • Software applications and systems, mobile applications, user interfaces
Consumer Goods and Retail
  • Anti-counterfeit devices
  • Beauty
  • Fashion
  • Health, fitness and sport
  • Household goods
  • Luxury goods
  • Moda
  • Retail
Electronics and Electrical Devices
  • Computer and IT architecture and system design
  • Medical devices
  • Optics
  • Restaurants and bars
  • Robotics
  • Scientific instruments
  • Semiconductor devices
  • Testing systems, control systems, signal processing
  • Travel and leisure
  • Wearable tech and human interfaces
Energy and Green Technologies
  • Biofuels
  • Clean air
  • Fuel cells and battery technology
  • Renewables and recycling
  • Solar power
  • Water, oil and gas, nuclear, fusion, fission
  • Wind turbines
Food and Beverage
  • Beverages
  • Nutrition
  • Retail
Industrial Manufacturing and Processing
  • Packaging
  • Polymers and plastics
  • Printing tech
  • Additive manufacturing
  • Coatings
  • Construction
Materials
  • Advanced materials
  • Composite materials
  • Metallurgy and alloys
  • Nanotechnology
Medical Devices and Diagnostics
  • Artificial intelligence
  • Devices
  • Diagnostic instruments
  • Digital health
  • Drug delivery
  • Immuno-oncology
  • In vitro diagnostics
  • Medtech
Pharmaceuticals
  • Diagnostics
  • Drug delivery
  • Generic market entry
  • Medicinal chemistry
  • Methods of production and synthesis
  • Personalised medicine/disease biomarkers
  • Pharmaceutical formulations
  • Pharmaceuticalsceutical formulations
  • Polymorphs
  • Small molecule pharmaceuticals
  • Supplementary protection certificates (SPCs)
Trade Marks
  • Consumer goods
  • Cosmetics and perfumery
  • Financial services
  • Mechanical engineering
  • Mechanical products
  • Services
  • Telecommunications