FDA’s GRAS Paradigm Shift: Is Mandatory Notification Lawful Implementation or Regulation by Fiat?

FDA has proposed transforming the voluntary notification program for Generally Recognized As Safe (GRAS) ingredients into a nationwide filing mandate. The proposal promises transparency but leaves foundational questions about statutory authority, scope, ingredient equivalence, and enforcement unresolved.

The proposal is particularly consequential for suppliers and companies using botanicals and other supplement-associated ingredients in conventional foods, including functional beverages, nutrition bars, and enhanced waters.

“Regulatory fiat” is provocative and should be used carefully. FDA is not claiming power from nowhere. The agency cites sections 201, 402, 409, and 701 of the Federal Food, Drug, and Cosmetic Act and offers a legal theory that mandatory GRAS notification would facilitate enforcement.

The question is whether that theory authorizes a prospective federal filing duty for persons relying on a statutory exclusion Congress defined by scientific status rather than agency notification.

That is a paradigm shift.

Subscribe to our newsletter to stay up to date with all our industry coverage.

Congress Defined GRAS by Scientific Status, Not Filing

Under section 201(s), a substance is excluded from the definition of “food additive” when it is generally recognized among qualified experts as having been adequately shown through scientific procedures to be safe under its intended conditions of use. Section 409 requires premarket authorization for food additives; GRAS substances are excluded from that requirement. The statutory distinction turns on safety status and intended use, not on whether a company submitted a GRAS notice.

Congress expressly created notification or registration requirements elsewhere in the Act. Section 409(h) establishes food contact notification; section 413 requires notice for certain new dietary ingredients; and section 415 requires food-facility registration. Congress did not add a notification requirement to the GRAS provision, including when the FDA Modernization Act of 1997 addressed food contact substances but left the GRAS exclusion untouched.

That omission does not automatically defeat FDA’s proposal, but it means FDA must demonstrate that mandatory notification legitimately implements existing authority rather than adding a condition Congress chose not to impose.

FDA’s strongest argument is section 701(a), which authorizes regulations for the “efficient enforcement” of the Act. FDA reasons that sections 409(a) and 409(d) authorize it to identify substances that may be food additives requiring approval, and that mandatory notices would reveal substances and uses the agency otherwise may not know exist.

That is a serious argument. FDA may investigate a marketed substance, determine an intended use is not GRAS, and pursue remedies against food containing an unsafe, unapproved additive, but the disputed step is whether authority also permits FDA to compel every covered person to make a prospective scientific submission.

FDA’s history makes that theory rather vulnerable. For decades, FDA described the statute as permitting manufacturers to make independent GRAS determinations and market substances without informing the agency. In 1997, FDA said a manufacturer could market a substance it determined to be GRAS “without informing the agency” and called that ability the manufacturer’s “prerogative under the statute.”

In the 2016 final GRAS rule, FDA said it lacked “express statutory authority” to require GRAS notices. In Center for Food Safety v. Becerra (2021), the government likewise argued that the statute “does not impose mandatory GRAS notification.” But the 2016 rule also acknowledged that a mandate might fall within FDA’s implied authority and would require separate notice-and-comment rulemaking — the caveat on which the current proposal relies.

FDA therefore disclaimed express authority, not necessarily implied authority. Still, it is reversing a voluntary policy maintained for decades, finalized in 2016, and later defended in court. It must explain why the same statutory structure now supports the opposite regulatory choice.

After Loper Bright (2024), a reviewing court would exercise independent judgment about the statute’s best reading rather than defer merely because FDA’s interpretation is reasonable. The proposal is not necessarily unlawful, but its implied-authority foundation faces closer judicial scrutiny.

A Mandate That Is Not Premarket Approval

The proposal appears designed to preserve the distinction between GRAS status and food-additive approval. FDA says the rule would not establish premarket review. A company could market before filing and during FDA review if the use actually is GRAS.

The filing duty is satisfied when FDA files the notice, not when it agrees with the conclusion.

Non-notification would not itself alter GRAS status or trigger a stop-sale order or recall. FDA says it would affect post-market review priorities.

That structure helps FDA argue it has not created an unauthorized approval program. Yet if filing is not approval and non-filing has no automatic market consequence, the mandate’s immediate effect is chiefly a new reporting obligation and information source.

Commercial Finality Without Legal Finality

The proposal’s most consequential procedural omission may be what happens when FDA disagrees with a notifier.

FDA could issue a public insufficient-basis letter after review. The company would have satisfied the notification requirement because FDA filed the notice, but that procedural compliance would offer little comfort.

FDA says an insufficient-basis letter would be relevant to whether the substance is an unapproved food additive, inform post-market action, and be public. For certain legacy uses, it would also foreclose the streamlined pathway. Customers, retailers, manufacturers, insurers, and plaintiffs could react immediately, before FDA formally determines the use is unlawful and before any court reviews that conclusion.

The final rule itself would be reviewable final agency action. A later insufficient-basis letter is harder. FDA could argue it is not final because it does not itself declare the ingredient unlawful, prohibit marketing, or compel a recall; the company could continue selling, submit another notice, or file a food-additive petition.

That position resembles FDA warning letters, which courts generally regard as advisory rather than final agency action because they do not themselves determine legal rights or compel compliance.

An insufficient-basis GRAS letter may be more consequential. It would follow a mandatory review that could last approximately 405 days, publicly state FDA’s adverse scientific position, inform post-market action, and bar eligible legacy uses from the streamlined pathway. Those features may support reviewability, but the proposal leaves the notifier to learn FDA’s answer when the market does.

The contrast with notification systems Congress expressly created is striking.

For new dietary ingredients, Congress authorized a petition for an order establishing conditions under which an NDI would reasonably be expected to be safe, required a decision within 180 days, and made the decision final agency action. For food contact notifications, the statute expressly makes an FDA objection final agency action subject to judicial review.

The proposed GRAS system has no comparable protection. FDA would create the filing obligation, conduct the scientific review, issue an adverse public response, and potentially characterize that response as less than final. That is both a fairness problem and a threat to FDA’s stated objective.

It also creates adverse selection. A company with a strong dossier may hesitate to file if an adverse letter could destroy its market without a reviewable decision; a company with a questionable ingredient has even more incentive to remain silent when non-notification merely increases post-market review priority. The system may thus discourage the submissions FDA most needs.

Before finalizing the rule, FDA should state whether an insufficient-basis letter is final agency action. If not, it should provide:

• An opportunity to cure deficiencies and seek reconsideration before publication of an adverse letter;
• A defined route and deadline for obtaining a reviewable final determination, with appropriate stay procedures; and
• Database language and interim protections clarifying that an insufficient-basis letter is not an adjudication that the substance is unsafe or unlawful.

FDA should also explain whether it would rely on such a letter alone for warning letters, import detention, removal from commerce, or judicial enforcement.

The rule should not make an adverse letter commercially final enough to eliminate an ingredient, but legally tentative enough to evade judicial review.

Information Is Not Enforcement

FDA’s stated purpose is to close an information gap, which is persuasive when the agency does not know a substance has entered the food supply, but less so when missing information is not the barrier to action.

Mandatory self-reporting is most likely to capture companies already inclined to comply. Established suppliers and brands will inventory uses, reconstruct dossiers, and submit notices — at FDA’s estimated 180 hours per full notice. Firms with questionable ingredients and weak safety support have the strongest incentive to remain silent.

FDA cannot treat non-notification as a prioritization factor until it discovers the company, substance, conditions of use, and absence of a qualifying notice. By then, surveillance or investigation has already overcome much of the information gap.

The proposal should therefore distinguish three problems:

• An information gap: FDA does not know a substance or use exists.
• A compliance gap: FDA knows about the use, but the company does not notify.
• An enforcement gap: FDA knows the substance, has reached an adverse position, but does not pursue compulsory domestic market action.

Mandatory notification addresses the first problem only when companies comply. For non-reporting firms, the information gap persists until FDA independently discovers the use; then the issue becomes compliance and potentially enforcement. The proposal creates a basis for acting against non-notification, but supplies neither surveillance to detect every unreported use nor compulsory action to remove products. That distinction matters because FDA cites substances it already knew about and had assessed adversely; warning letters and import alerts are not seizures, injunctions, mandatory recalls, or other compulsory domestic action.

The question is not whether more information has value, but how more notices will change enforcement outcomes when information was not the limiting factor in several examples FDA cites.

One Notice Does Not Necessarily Mean the Same Substance

Proposed section 170.205(a) applies to “any person” introducing a substance into interstate commerce under the GRAS provision, without clearly identifying which supply-chain party must file. If each person must file, FDA’s burden estimates may omit downstream marketers. If one notice can cover the supply chain, the rule must say when another company may rely on it.

That reliance question cannot be answered by ingredient names alone. A one-notice model may work for a discrete chemical made to equivalent specifications, but botanical extracts can vary materially by species, plant part, source, extraction process, ratio, standardization, constituent profile, contaminants, residual solvents, and physical form.

Even two turmeric extracts standardized to 95% curcuminoids may not be toxicologically or compositionally equivalent if their source material or extraction processes differ. Safety may depend on the full profile of the tested material, not merely the named marker.

FDA recognizes that a no-questions letter may not cover another material when identity, manufacturing, specifications, impurities, physical form, food categories, use levels, or technical effect differ significantly. Yet the proposal neither defines “significantly different” nor provides a process for showing one supplier’s botanical falls within another’s notice.

The final rule needs a defined-material approach: one notice per substance, manufacturing envelope, specifications, and conditions of use; objective comparability and change-control criteria; and a master-file or right-of-reference mechanism protecting confidential information while letting downstream companies confirm coverage.

FDA also proposes a one-year streamlined pathway for certain existing uses. A qualifying submission could enter a public list without the full safety dossier, and others could rely on the listing for the same conditions of use.

But identical conditions of use do not establish identical ingredients. A botanical name and use level could make unlike extracts appear covered despite material differences in composition or manufacturing.

FDA stresses that listing would not be a safety determination and could later require a full notice or food-additive petition. But a public database is useful only if each entry defines the material it actually covers.

The proposal may reveal good-faith GRAS conclusions and prompt review of old dossiers, but transparency is not safety review or enforcement. If the immediate problem is marketplace visibility, another mechanism may fit better.

Before finalizing the rule, FDA should identify the statutory delegation; designate the notifier; establish cross-company reliance rules; define ingredient comparability; explain the consequence of non-notification; and show how the information will produce public-health action.

Affected suppliers and conventional-food companies should ask:

• Who authorized the mandate?
• Who must comply?
• What counts as the same substance?
• What follows from non-notification?
• How will more information produce more enforcement?

An Inventory-First Path to Market Visibility

FDA maintains the Substances Added to Food (SAF) database, which it describes as only a partial list of food ingredients. Because companies may market substances based on independent GRAS conclusions without notifying FDA, the agency says it lacks a complete picture of substances and uses entering the market.

The proposal seeks a “more comprehensive catalog” plus supporting information. But if the immediate problem is identifying what is used, requiring a full GRAS notice for new uses may be more burdensome than necessary.

If FDA concludes section 701(a) authorizes mandatory reporting for purported GRAS uses, an inventory-first rule could achieve much of the transparency objective without requiring every scientific rationale. FDA could require persons marketing substances not covered by an existing regulatory pathway or no-questions letter to report the substance name, conditions of use, and food categories, creating a comprehensive inventory.

That approach would separate information gathering from scientific evaluation. FDA would first identify marketed substances, then use post-market review to target safety concerns warranting enforcement, a full GRAS notice, or a food-additive petition. Industry would face a registration-style submission rather than a 180-hour dossier, reducing cost and disclosure risk. FDA could also accept verified third-party nominations as investigatory leads.

FDA could cross-reference the SAF database, GRAS notice inventory, food additive regulations, and commercial ingredient databases to build a preliminary list. Substances unaccounted for through existing pathways would form the residual population FDA seeks. Targeted review and enforcement could change behavior more directly than a universal filing mandate whose main consequence for non-compliance is priority for possible post-market review.

FDA invites comment on alternatives. An inventory-first, enforce-second approach may achieve transparency at lower cost and with broader compliance. FDA could begin assembling the database from existing resources while developing the framework through notice-and-comment rulemaking. The question is why it has not already done so.

Conclusion

FDA has advanced an implied-authority argument under section 701(a), reserved in 2016. But Congress defined GRAS through scientific status, not agency filing, and expressly created notification systems elsewhere.

The proposal leaves unresolved who must file, when one company may rely on another’s notice, what counts as the same substance, whether an insufficient-basis letter is reviewable final agency action, what follows from non-notification, and how more information will produce more enforcement.

Those issues bear directly on the proposal’s statutory fit, economic analysis, fairness, and likelihood of accomplishing its public-health purpose.

Whether the proposal is lawful implementation or regulation by fiat will depend on the statute’s best reading, the administrative record, and FDA’s response to these practical limitations.

A paradigm shift requires more than a new form. It requires clear authority, administrable boundaries, and a credible account of what the regulation will actually accomplish.

Comments are due Dec. 9, 2026. FDA has stated it will not consider late submissions.