> ## Content Index
> Fetch the complete content index at: https://www.fdaweb.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# FDA Mulls Artificial Intelligence Use in Drug Manufacturing
- URL: https://www.fdaweb.com/fda-mulls-artificial-intelligence-use-in-drug-manufacturing/
- Published: 2023-03-01T12:00:00.000Z
- Updated: 2026-09-14T18:17:47.000Z
- Author: David McFarland
- Tags: Drugs, #legacy-id-D5153876

FDA has outlined new potential GMP issues under policy consideration in a [discussion paper ](https://www.fda.gov/media/165743/download?ref=fdaweb.com)entitled “Artificial Intelligence (AI) in Drug Manufacturing.” The agency says difficulties could result from “potential ambiguity” on how to apply existing GMP regulations to AI or due to a lack of agency guidance or experience. The considerations discussed are open for stakeholder comments as FDA works through the issues.

The agency says one area under consideration involves cloud applications that may affect oversight of pharmaceutical manufacturing data and records. “Third-party data management systems could be used for functions that extend beyond data storage,” the paper says. “For example, data stored in these systems may be analyzed by AI to support models for process monitoring and \[advanced process controls\] APC. Data integrity and data quality must be ensured in these environments.”

FDA points out that existing quality agreements between a manufacturer and a third party (e.g., for cloud data management) may have “gaps” related to AI risk management in the context of manufacturing monitoring and control. “During inspections, this may lead to challenges in ensuring that the third party creates and updates AI software with appropriate safeguards for data safety and security,” it says. “Further, FDA inspection approaches for evidence gathering of records management may need to expand due to the complexity of managing third-party cloud data and models.” To illustrate this concern, the agency says that ongoing interactions between cloud applications and process controls could “complicate” establishing data traceability, and could create potential cybersecurity vulnerabilities, thus requiring an assessment of procedures in place to monitor data integrity vulnerabilities during an inspection.

Another area under consideration involves digitization of manufacturing controls that likely will generate more information about a process and product, including increased data collection frequency and more types of data recorded. FDA says that if the “raw data collected during the manufacturing process increases significantly, there may be a need to balance data integrity and retention with the logistics of data management. Applicants may need clarity regarding regulatory compliance for generated data (e.g., which data needs to be stored and/or reviewed and how loss of these data would impact future quality decisions such as product recalls). Further, applicants may need additional clarity for data sampling rates, data compression, or other data management approaches to ensure that an accurate record of the drug manufacturing process is maintained.”

Other areas in the paper under consideration and open for comments include:

- Applicants may need clarity about whether and how the AI use in pharmaceutical manufacturing is subject to regulatory oversight.
- Applicants may need standards for developing and validating AI models used for process control and to support release testing.
- Continuously learning AI systems that adapt to real-time data may challenge regulatory assessment and oversight.