Skip links

Artificial Intelligence (AI), Machine Learning (ML) and the FDA

“Artificial Intelligence (AI) and Machine Learning (ML) as a branch of computer science, statistics, and engineering that uses algorithms or models to perform tasks and exhibit behaviors such as learning, making decisions, and making predictions. ML is considered a subset of AI that allows models to be developed by training algorithms through analysis of data, without models being explicitly programmed.” – (FDA, 2023)

Since the FDA approved its first AI/ML-enabled medical device in 1995, the agency has been committed to updating and evolving the regulatory landscape to accommodate this new and rapidly changing field. In 2019, the FDA published a discussion paper on a proposed regulatory framework for modification to AI/ML-based software as a medical device (SaMD), which received a significant amount of feedback from stakeholders. This feedback prompted the FDA to release a second paper in 2021 titled “Artificial Intelligence/Machine Learning (AI.ML)-Based Software as a Medical Device (SaMD) Action Plan,” which outlined the agency’s five-pillar approach to regulating AI/ML. Additionally, the FDA’s Center for Drug Evaluation and Research (CDER) released a discussion paper this year to facilitate early feedback from stakeholders outside the Agency to consider when developing a future AI/ML regulatory framework.

FDA Discussions

As a result of the feedback received in 2019, the FDA outlined the five-pillar approach to regulating AI/ML. The five pillars identified are:

  1. Tailored Regulatory Framework for AI/ML-Based SaMD
    • The FDA realizes the need for AI/ML regulatory guidance, support, and oversight of AI/ML-enabled devices.
  2. Good Machine Learning Practices (GMLP) Development
    • The FDA is committed to working with domestic and international regulatory communities and agencies to create a set of international standards and development maintenance of AI/ML-enabled medical devices. These standards will help guide the development and maintenance of these devices.
  3. Patient-Centered Approach Incorporating Transparency to Users
    • The FDA stressed the need and commitment for creating forums to disseminate information about AI/ML-enabled devices to the general public. This information includes the source of data used in algorithm training and information about the benefits, risks, and limitations of the devices.
  4. Regulatory Science Methods Related to Algorithm Bias and Robustness
    • The FDA acknowledges that there are unique concerns related to bias and generalizability of use associated with AI/ML-enabled devices and the data forming the basis for their functions. The FDA is committed to supporting regulatory science research methods to ensure that factors such as race, ethnicity, and socioeconomic status are considered in the development and maintenance of AI/ML-enabled devices.
  5. Incorporation of Real-World Performance Metrics
    • The FDA recognizes that real-world data is particularly important for the development and regulation of AI/ML-based medical devices as these devices continuously learn and adapt based on real-world data. The FDA is committed to piloting real-world performance monitoring on a voluntary basis to develop uniform real-world performance principles applicable to AI/ML-based device oversight.

(FDA, 2021)

This year, FDA’s CDER released a discussion paper to facilitate early feedback from stakeholders outside the Agency to consider when developing a future AI/ML regulatory framework. In the discussion paper, CDER listed potential applications of AI/ML-based devices to be used in the manufacturing process of products. They included:

  1. Process Design & Scale-Up
    • Utilizing AI to optimizes processing development data and reducing development time & waste.
  2. Advanced Process Control (APC)
    • AI can be used to assist developing process control that can help predict the progression of a process.
  3. Process Monitoring and Fault Detection
    • AI can be used to assist in monitoring equipment and help detect changes from normal performance. AI can help monitor product quality and detect deviations.
  4. Trend Monitoring
    • AI can assist in examining consumer complaints and reports that contain large volumes of text. AI will be able to identify trends in manufacturing-related deviations.

(FDA, 2023)

These examples are not exhaustive and the potential applications of AI in pharmaceutical manufacturing may continue evolving.

There have been multiple guidance documents aimed at providing industry oversight of regulatory activities related to AI/ML-enabled devices since the release of these action plan. All the FDA guidance documents can be found on the FDA website at

FDA Support for AI/ML Drug Development

The use of artificial intelligence (AI) and machine learning (ML) in drug development has increased rapidly, affecting areas directly overseen by the FDA, such as clinical trial design, digital health technologies (DHTs), and real-world data (RWD) analytics. The number of drug submissions involving AI/ML has also increased.

The FDA is actively supporting this increased AI/ML integration in a variety of ways, including:

  • Establishing the CDER AI Steering Committee (AISC) to provide guidance and oversight on AI/ML in drug development.
  • Providing consultations for drug submissions that involve AI/ML-based devices to help ensure that these submissions meet regulatory requirements.
  • Developing a framework for AI/ML-based devices to ensure that these devices are safe and effective.
  • Creating workshops and holding meetings to provide additional support to stakeholders on AI/ML in drug development.

(FDA, 2023)

As the use of AI/ML in drug development expands, so will the regulatory requirements for these technologies. These requirements will continue to evolve as AI/ML is applied to new areas of the drug development process.


The introduction of AI and ML in the pharmaceutical industry has marked a significant milestone in the evolution of regulatory and manufacturing processes. The FDA’s commitment to AI and ML is evident in the publication of discussion papers, development of regulatory frameworks, and the engagement of stakeholders to gather valuable feedback. Premier Consulting can assist you in navigating the complexities of AI and ML regulatory requirements for your products. Our team can provide guidance, insights, and solutions to ensure compliance with evolving regulatory requirements. Contact us today to find out how we can support your program.


June Morrison, Regulatory Operations Associate II

Jorge Sierra, Regulatory Operations Associate III