AI is becoming an increasingly important tool across many sectors, not least life sciences. AI is being deployed in many areas such as drug discovery, diagnostic data interpretation, app development. The use of AI, however, presents particular difficulties for the protection of AI-generated innovations and the use of material in their development. In this article we consider these issues.
Can the AI itself be an inventor?
Patent systems normally recognise one or more individuals as inventors of a patented invention. The initial ownership of the invention will rest with them or their employers. What happens when an invention is apparently generated by the AI itself?
The UK Supreme Court ruled in December 2023 on a case where a patent application named an AI machine (DABUS) as the patent inventor. The applicant, Dr Stephen Thaler, maintained that DABUS had been acting autonomously and that he had acquired rights to the invention though his ownership of it.
Dr Thaler has been testing the AI-as-inventor concept in patent systems around the world. In Australia, an initial decision to recognise an AI as inventor was eventually overruled. The Federal Court analysed the concept of inventor in detail, concluding that case law showed the grant of a patent was “premised upon an invention for the purposes of the Patents Act arising from the mind of a natural person or persons”.
At the European Patent Office, the machine-as-inventor concept was rejected, both by the patent examiner and the Board of Appeal. The Board commented that the user or owner of the device involved in an inventive activity could designate themselves as inventor under the current law. If any new approaches were needed, it would be for lawmakers to consider changes to the European Patent Convention.
The UK Supreme Court has taken a similar path, deciding that a patent inventor must be a natural person. However, the court also concluded that Dr Thaler could not, as owner of DABUS, nevertheless apply for a patent himself. Patent law “does not confer on any person a right to obtain a patent for any new product or process created or generated autonomously by a machine…let alone a person who claims that right purely on the basis of ownership of the machine”. Allowing for this kind of ownership right would require a change in the law.
Is an AI-related invention within the “computer program” exclusion?
Patent law generally excludes computer programs from protection. These receive protection under copyright law instead. Does this mean that AI engines themselves should be excluded from protection?
This issue was examined in a case about patent protection for an artificial neural network (ANN) for providing media file recommendations to an end user. Patent protection was initially refused. However, an appeal to the High Court saw the decision reversed. Analysing the questions; “what is the computer?”, and “where is the program which is said to engage the exclusion?” led the judge to conclude that it was “appropriate to look at the emulated ANN as, in substance, operating at a different level (albeit metaphorically) from the underlying software on the computer, and it is operating in the same way as the hardware ANN”. The claim was not to a “computer program” and so the exclusion did not apply.
This case is discussed in more detail here. It leaves the door open to claims around new AI approaches to problems, provided that the thresholds of novelty and inventive step can be reached.
Is there protection for the data used in AI development?
Databases can benefit from certain kinds of protection, sometimes through copyright or database protection. These forms of protection do have limitations, however. Copying must usually be proved, and there are often exclusions from protection for research activity.
Maintaining confidentiality around the data can also offer effective protection, although great care must then be taken to limit access and control publication of content.
Litigation over the use of publicly available material is currently ongoing in New York, with the New York Times bringing a copyright infringement lawsuit against Microsoft and OpenAI. The New York Times argues that extensive use of its material in building the defendants’ AI systems is unlawful without permission or payment from the copyright owners. The claim is based not only on use as training data, but also on the replication of input material in outputs. The dispute may well be settled in the context of ongoing negotiations, although developments will be closely watched as it unfolds.
This dispute highlights the current legal tensions between those holding database resources and those seeking to mine them to develop new insights and innovations.
How does this map onto life sciences?
These legal uncertainties need to weighed by both database owners and AI developers in the life sciences arena. While huge progress may result from deploying data resources in new ways, finding the right balance of ownership and recompense is not easy. We see collaborators working together to bring new insights for the benefit of patients and to find cost-effective solutions to problems. Looking forward to what the results might be and how to control and share them effectively is key to forming successful partnerships.
Learn more about our intellectual property and life sciences services.
Our content explained
Every piece of content we create is correct on the date it’s published but please don’t rely on it as legal advice. If you’d like to speak to us about your own legal requirements, please contact one of our expert lawyers.