Artificial Intelligence (AI) is transforming the life-sciences industry by making discoveries from massive biological data using machine learning, integrating clinical records and genomic data of different kinds, discovering new medicine or drug targets, identifying new classes of cell types, carrying out diagnostics, or customizing clinical procedures in precision medicine.

Artificial Intelligence involves a number of different technologies, primarily machine learning, deep learning, neural networks, natural language processing, and computer vision. There is a considerable degree of connection among them, but the core technology is machine learning. So, bearing in mind that this technology could fall in the exception of patentability under some legislation, it is necessary to consider new regulations to protect the IP rights. Thus, several aspects should be considered when protecting this technology and choosing how to protect it.

Fields of applications in life sciences

Before continuing, it is important to briefly explain what is Artificial Intelligence. It is defined as computer systems able to perform tasks normally requiring human intelligence such as visual perception, speech recognition, decision-making, and translation between languages.

Artificial Intelligence originates from computer science and covers a wide range of approaches intended to enhance the ability of machines to make data-driven decisions and accurate predictions of events. In many scientific fields, AI is being increasingly considered and integrated, especially in the context of Big Data. Given their complexity and highly interdisciplinary nature, life sciences provide ample opportunities for AI to impact R&D efforts in a variety of ways.

There are numerous areas where the life-science industry uses AI effectively today. Some of them are the following:

  • Pharmaceutical research, pharmacology, and drug discovery
  • Accelerating Drug Development

Scientists are integrating research data, lab data, and clinical data, in combination with new information sources (e.g., social media and wearables) across the drug development spectrum, creating a holistic picture of the drug development candidate. Improving ways to acquire and mine data in real time allows scientists to use AI and machine learning to make better decisions faster, which will accelerate the product development and scale-up process.

  • Epidemiology and clinical investigations
  • In silico modeling and simulation of molecular systems and organisms
  • Designing Clinical Trials

Artificial Intelligence can design clinical trials, estimating the ideal sample size, and implementing them remotely on participants across a wider geographical area. This, in turn, reduces the cost and increases the odds of obtaining relevant and accurate data.

  • Introducing Robotic Surgery

Robotic surgery is a new field that is garnering a considerable amount of interest. Nowadays, surgeries can be performed in previously inaccessible places. Once trained, a robot will be competent enough to perform each operation consistently and accurately. The consistency and accuracy of the surgery will be irrespective of the duration of the surgery. It is touted to be superior when compared to human performance, which will predictably decline with time.

  • Developing the next-generation of radiology tools

The current diagnostics processes rely on either invasive techniques or information gathered from radiological images. These include data from CT scans, X-rays, or MRI machines. AI-based radiology tools will enable clinicians to gain a more precise and detailed understanding of how a disease progresses by performing virtual biopsies.

  • Telemedicine

Unavailability or dearth of trained professionals such as radiologists or ultrasound technicians can considerably limit access to life-saving care. This is commonly observed in emergent and developing parts of the world. The AI-powered tool “Telemedicine,” which equips patients to tackle and prevent certain health concerns, has become popular in such regions. The health care start-up “WeDoctor” can independently conduct eleven diagnostic tests and upload data for consultation in an automated fashion.

  • Clinical Trials

Clinical Internet of Things refers to the ability of patients to wear mobile devices and sensors that will capture and provide a stream of quality, nearly real-time data to researchers. AI is the technology those researchers will use to analyze the data and look for information, insights, or patterns. It has been defined as machines being able to perform “smart” tasks that are characteristic of human intelligence. Machine learning is a term that refers to the ability of AI algorithms to learn and develop without being explicitly programmed.

Life sciences companies are likely to begin experimenting further with AI in their workflows in the coming years, but they face challenges in AI adoption due to strict regulations.

The regulatory challenge

Artificial Intelligence and Machine Learning (AI/ML) involves new computing technologies, and vast amounts of training data that pose new regulatory challenges such as:

  • Determining the accountability for providers of AI/ML-based solutions and assigning liability for harm caused by the “black box” of an AI/ML process.
  • Assuring the quality and safety of products or therapies developed using AI/ML.
  • Guaranteeing that the data and/or algorithms used in AI/ML solutions are not biased against underserved populations.
  • Ensuring that the AI/ML solutions developed are sustainable and environmentally friendly, and that the recommendations are not in conflict with societal priorities such as social justice.
  • Protecting the privacy of patients and their data. Existing patient privacy rules do not, for example, protect patient data when it is shared and used by technical or consumer marketing organizations rather than healthcare providers.

Thus, international and national legislations must be adapted or must be created to regulate the safe use and protection of the AI-related IP rights.

Protection of Artificial Intelligence Innovations in Life Sciences

A substantial investment in building and deploying machine learning (ML) technology—the most active area of AI technology being developed today—warrants carefully considering how to protect the resulting intellectual property rights, but there are challenges in doing so. Several aspects should be considered when protecting this technology and choosing how it is to be protected, which would be with a patent o with trade secret protection.

Trade Secret Protection

Protecting by Trade Secret, there is no time limit on trade secret protection so long as the subject matter is kept secret, and there are no eligibility, novelty, or obviousness bars to clear. There is, however, no recourse for independent discovery by a competitor. Important factors to consider when weighing trade secret and patent protection include:

1. Detectability. If detecting when a competitor uses an invention is hard, then the value of patenting that invention is diminished because it will be difficult to know that the patent is being infringed. This may be the case with innovative training algorithms for ML systems—it is perhaps possible to detect that the ML system is being used, but hard to detect how it was trained. This might suggest the trade secret route for such technology.

2. Reverse Engineering. If it is easy to reverse engineer the invention or hard to keep it secret (e.g., due to desire to publish or visibility of the invention in the product), then the patent route may be preferable.

Trade secrets offer a degree of protection in circumstances where patenting is not the best approach.

Keeping part of an invention secret is an option if:

• time is needed to generate more experimental data to ensure optimal scope of protection;

• the invention could not be described in a reproducible way without disclosing training data that should remain secret;

• patent case law is not favorable in terms of patent eligibility;

• infringement is hard to detect;

• the lifecycle of the invention is short; and

• the filing behavior of the competition is not active.

Trade secret protection can be very cost-effective since there are no official fees to pay. However, there are management and administrative costs to businesses since comprehensive policies and procedures are needed to track and secure trade secrets.

Trade secrets offer a degree of protection in circumstances where patenting is not the best approach.

If the technology needs to be known by several entities, such as software contractors, customers, and a large number of employees, then it may not be practical to be kept secret and trade secret protection is not suitable.

Protection by Patent

These inventions must comply also with the requirements of novelty, inventive step, and industrial applicability to be patented. If claims relate to a method involving the use of technical means, for instance a computer or a device, the subject matter in its entirety is of a technical nature and is patentable as an invention. The question, then, is whether the invention satisfies other requirements of patentability, in particular novelty and inventive step.

The evaluation of the inventive step, widely considered the more problematic requirement, assesses whether the mathematical method contributes to producing a technical effect that serves a technical purpose. For example, an X-ray apparatus providing a genotype estimate based on an analysis of DNA samples or an automated system providing a medical diagnosis by processing physiological measurements.

Some examples of potentially patentable aspects of an ML system are:

  • New ML model. In deploying ML technology, a new model may have been developed (e.g., new neural network architecture). Claiming novel aspects of the model will help to address novelty and inventive step challenges.
  • Training an ML model. Innovative ways of generating training data and/or a new training algorithm may be claimed. For example, when there is insufficient training data, it may be augmented by synthesizing new training data from old training data or other sources, and such data augmentation techniques may be innovative and, thus, the focus of patent claims.
  • Organizing an ML model. How an ML model is integrated into an application may provide a novelty and inventive step hook. Claims focusing on integration and deployment should go beyond merely displaying the model’s output and focus on what the output is used to achieve. For example, applying an ML system to medical images may result in instructions to take more images with different settings because the ones obtained are unsatisfactory. Other examples include choosing among different next steps in a control system, customizing a patient’s treatment, or updating a clinical trial. When an ML system is deployed in conjunction with a specialized device (e.g., an imaging device, a sequencing device), rather than merely a computer, claims could focus on how the ML system is integrated with the device.
  • By publishing a patent application about an AI algorithm for finding new drugs, there is a possibility that it is harder to gain patent protection for individual new drugs found using the AI algorithm. This is because the new drugs are arguably obvious since the AI algorithm is known.

Filing of AI Patent applications

The number of AI-based patent filings has increased rapidly in recent years, particularly in the United States and Asia. Even in Europe, patent filings grew at an annualized rate of over 50% from 2014 to 2017.

Machine learning is the dominant AI technique disclosed in patents. Nevertheless, according to the field of application the main fields are the following:

  • Transportation industries (15 percent of all AI-related patents),
  • Telecommunications (15 percent), and
  • Life and medical sciences (12 percent).
  • Transportation, agriculture, and computing in government are growing industries, with annual growth rates of at least 30 percent between 2013 and 2016.

Patent families for application field categories and sub-categories

Graphic 1 Uhthoff

Overall number of patent applications by Patent Office

The greatest number of patent applications are filed in the patent offices of U.S. and China, followed by Japan, while WIPO and the EPO are also often used.

Graphic 2 Uhthoff


AI is expected to revolutionize processes across a wide range of fields. It is foreseen that AI will also affect intellectual property rights, in particular patent rights and their management. This is likely to be a two-way process: on the one hand, AI developments will affect and be incorporated into IP rights management; on the other hand, IP policies and practices will interact with the strategy of managing innovation in AI.

In addition, as AI develops, some of the questions that are currently discussed only hypothetically may become real issues. These include the inventorship of AI, patent- and more generally IP-rights infringement by AI. Such questions may call for related regulation or a certain interpretation of existing regulations to cover possible gaps and answer related questions.

Janett Lumbreras



Ver también

Mujeres Abogadas

En el transcurso de la historia, las mujeres han luchado incansablemente por su libertad, derechos e igualdad. Lamentablemente, como sociedad, hemos normalizado tanto el machismo y


The industrial and economic development that globally has occurred since the last century, has generated an important impact not only related to companies’ commercial growth, but