FDA GRAS Reform moves forward in Federal Rulemaking Agenda

‍FDA has advanced a rulemaking on GRAS Notification requirements for substances used in human and animal food. The proposed rule is included in the latest federal Unified Agenda under RIN 0910-AJ02 and is listed for publication in December 2026. If finalized, companies would be required to submit GRAS notices to FDA for certain substance uses. Under current regulations, submission of a GRAS notice is voluntary, and companies may self-affirm that a substance use is GRAS without notifying FDA. ‍‍ ‍

The proposal follows earlier federal action on the self-affirmed GRAS pathway. In March 2025, the U.S. Department of Health and Human Services (HHS) directed FDA to explore rulemaking to revise the GRAS framework and eliminate the self-affirmed GRAS pathway. ‍‍ ‍

The text of the proposed rule is not yet public. Details remain unknown, including which GRAS uses would require notification, whether exemptions would apply, and how existing self-affirmed GRAS conclusions would be handled.‍‍ ‍

The proposed change may affect companies that rely on self-affirmed GRAS conclusions. For certain ingredient uses subject to mandatory notification, an internal self-affirmed GRAS conclusion would no longer be sufficient. Therefore, companies that have used self-affirmed GRAS, or plan to use this pathway, should prepare for the possibility that a GRAS notice will need to be submitted to FDA. This may require converting an internal GRAS assessment into a submission-ready GRAS notice or identifying data gaps before FDA notification is required. ‍

Vireo is following these developments closely. We have extensive experience supporting companies with GRAS strategies, data gap assessments, safety substantiation, and GRAS notice preparation. Please reach out if you would like to discuss what these developments may mean for your product or ingredient regulatory or commercialization strategy.

Next
Next

Vireo Advisors Contributes to AI-derived Open-Source Growth Factor Thermostability Dataset