FDA Shares Approach to Regulating AI: What Life Science Companies Need to Know

The integration of artificial intelligence (AI) and machine learning (ML) into health care and biomedicine is transforming the industry by enhancing diagnostics, treatment planning, and patient care. A recent article titled “FDA Perspective on the Regulation of Artificial Intelligence in Health Care and Biomedicine” in the Journal of the American Medical Association (JAMA), authored by Haider J. Warraich, MD, Cardiologist at Brigham and Women’s Hospital; Troy Tazbaz, BS, Director of Digital Health at US FDA; and Robert M. Califf, MD, Commissioner, US FDA, discusses the U.S. Food and Drug Administration’s (FDA) current approach to regulating AI and machine learning (ML) technologies.

FDA’s History Regulating AI-Enabled Products

FDA defines AI as a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments. Artificial intelligence systems use machine- and human-based inputs to perceive real and virtual environments; abstract such perceptions into models through analysis in an automated manner; and use model inference to formulate options for information or action

In 1995, the FDA approved PAPNET, a neural network software designed to reduce misdiagnosis of cervical cancer in Papanicolaou tests. Despite its higher accuracy compared to human pathologists, it wasn’t widely adopted due to cost-effectiveness issues. Since then, the FDA has authorized approximately 1,000 AI-enabled medical devices, predominantly used in radiology and cardiology.

The FDA has also seen a significant increase in regulatory submissions incorporating AI for drug discovery and development. In 2021, there were 132 applications—a tenfold increase from 2020. Today, AI applications include drug discovery, clinical trial design, dose optimization, adherence monitoring, endpoint assessment, and post-market surveillance, especially in oncology and mental health.

Evolving Regulatory Frameworks

This timely article specifically states, “The evolution of AI illustrates a major quality and regulatory dilemma. Since the safety and effectiveness of many AI models depends on recurrent evaluation of their operating characteristics, the scale of effort needed could be beyond any current regulatory scheme. For traditional medical products, which generally remain the same no matter where they are distributed, appropriately robust studies for a specific indication can provide assurance of safety and effectiveness of the product for the US population, but this may not necessarily be the case for many AI-enabled medical products given their contextual sensitivity. Therefore, this is an important area necessitating close scrutiny as AI becomes more routine.”

The FDA regulates 20% of the U.S. economy and must adapt to oversee AI’s safe and effective use while fostering innovation. In January 2021, the FDA released a five-point action plan to advance AI and ML-enabled software, committing to a tailored regulatory framework for AI-enabled medical devices. In 2022, the FDA published final guidance on clinical decision-support software, clarifying policies related to AI applications under the amended device definitions from the 21st Century Cures Act. Earlier this year, FDA also proposed an AI lifecycle management framework.

This year, FDA’s medical product centers identified four key focus areas for the development and use of AI in medical products:

Enhancing partnerships among stakeholders to ensure that AI technologies are used safely and effectively in healthcare.
Developing unified standards, guidelines, best practices, and tools to ensure consistency and reliability in AI applications.
Evolving regulatory frameworks to encourage innovation while maintaining safety and efficacy standards for AI-enabled products.
Investing in research to assess and continuously monitor the performance of AI technologies to ensure they meet healthcare needs.

Read Axendia’s full article on this topic here.

Use Cases for AI

According to FDA, the potential of AI to inform multiple aspects of medical product development is profound and already underway. Use cases include:

Drug Target Identification, Selection, and Prioritization
Screening and Designing Compounds
Modeling Pharmacokinetics and Pharmacodynamics
Clinical Trial Data Collection, Management, and Analysis
Postmarket Safety Surveillance and Evaluation
Advanced Pharmaceutical Manufacturing
Maintaining Robust Supply Chains
Anticipate or rapidly respond to shortages

Why Life Science Companies Should Take Notice

Regulatory Compliance: Understanding the FDA’s perspective is critical for life science companies developing AI/ML technologies. Compliance with regulatory requirements ensures a smoother approval process and reduces the risk of costly delays or rejections.

Product Development Strategies: Knowledge of the FDA’s approach allows companies to design their AI/ML products with regulatory considerations in mind from the outset, leading to more efficient development cycles.

Market Competitiveness: Companies that proactively align with FDA guidelines can gain a competitive advantage by bringing safe, effective, and trustworthy AI/ML products to market faster.

Risk Management: Addressing concerns such as data bias, algorithm transparency, and post-market monitoring reduces potential legal and ethical risks associated with AI/ML devices.

Innovation Encouragement: The FDA’s willingness to work with industry players opens opportunities for life science companies to influence future regulatory policies and to contribute to setting industry standards.

In Brief

The FDA’s perspective on regulating AI in healthcare and biomedicine highlights the balance between fostering innovation and ensuring patient safety. Strong oversight by the FDA protects the long-term success of industries by focusing on evaluation to advance regulated technologies that improve health. The FDA will continue to play a central role in ensuring safe, effective, and trustworthy AI tools to improve the lives of patients and clinicians alike. However, all involved entities will need to attend to AI with the rigor this transformative technology merits.

Life science companies should closely follow these regulatory developments to navigate the evolving landscape effectively. By integrating the FDA’s guidelines into their development processes, companies can not only achieve compliance but also enhance the quality and reliability of their AI/ML technologies, ultimately contributing to better healthcare outcomes.

To discuss how to best leverage AI in your organization, click on this link to schedule an Analyst Inquiry on this topic.