My Healthcare News & Research — August 10, 2026 · The GRAS Loophole: What the FDA’s New Rule Would Actually Change

2026 08 10 — scientific infographic poster

On August 10, 2026, the Department of Health and Human Services announced a proposed rule that would, for the first time in the sixty-eight years since Congress created the category, require food companies to tell the FDA when they decide on their own that an ingredient is safe. Health Secretary Robert F. Kennedy Jr. said the administration was closing the GRAS loophole “once and for all.”

It does not close it. We read the proposed rule — all 140 pages of the pre-publication text filed with the Federal Register — and the document says so itself, in plain language, on more than one page. What the rule does is real, and it is worth understanding. What it does not do is the part most of the coverage has skimmed past, and it is the part that determines whether anything in your grocery cart changes.

This article walks through both: the actual mechanism of the GRAS loophole, what the proposal would and would not require, why the enforcement provision is weaker than the word “mandatory” suggests, the legal challenge that is almost certainly coming, and the realistic timeline — which runs to about 2030 before a single label changes. There is also a section on how to file a public comment, because the comment window is open right now and closes December 9, 2026.

Table of Contents

  1. The Announcement at a Glance
  2. What GRAS Actually Is
  3. How a Grandfather Clause Became a Loophole
  4. How Big Is the Blind Spot?
  5. What the Proposed Rule Would Require
  6. The Streamlined Pathway for Ingredients Already on Shelves
  7. What the Rule Explicitly Does Not Do
  8. “Mandatory” Without a Penalty
  9. The Legal Question: Can FDA Do This Without Congress?
  10. The Timeline: Roughly 2030
  11. The Separate Ultra-Processed Food Announcement
  12. What Supporters and Critics Are Saying
  13. What This Means for You
  14. How to File a Public Comment
  15. Sources and Primary Documents
  16. Connections
  17. Featured Videos

The Announcement at a Glance

The formal document is a proposed rule titled “Substances Generally Recognized as Safe,” Docket No. FDA-2025-N-3262, RIN 0910-AJ02. It amends 21 CFR parts 170 (human food) and 570 (animal food). It was filed for public inspection on August 10 and scheduled for publication in the Federal Register on August 11, 2026.

Here is the whole thing in six lines:

An honest one-sentence summary: this is a disclosure rule, not a safety-approval rule. It builds the catalogue that has been missing since 1958. It does not give the FDA the power to say no.

What GRAS Actually Is

In 1958, Congress passed the Food Additives Amendment to the Federal Food, Drug, and Cosmetic Act. It did something genuinely important: it created a premarket approval system for food additives. Before you could add a new chemical to food, you had to prove to the FDA that it was safe, and the FDA had to agree.

But Congress immediately ran into an obvious practical problem. If “food additive” meant anything added to food, then salt, vinegar, black pepper, baking soda and mustard seed would all need federal safety dossiers. That was absurd. Humanity had been eating those for millennia.

So the 1958 amendment carved out an exception. A substance is not a “food additive” — and therefore does not need premarket approval — if it is generally recognized as safe among qualified experts, based on either a long history of use in food before 1958 or published scientific evidence. That is the GRAS category, defined in section 201(s) of the FD&C Act. Salt is GRAS. Vinegar is GRAS. This was a sensible piece of lawmaking.

The critical structural fact, and the one everything else follows from: GRAS is a self-executing legal status, not a permission slip. A substance is GRAS because the scientific consensus says it is, not because the FDA has blessed it. That means somebody has to decide whether the consensus exists. Congress never said who.

How a Grandfather Clause Became a Loophole

For a while, the FDA tried to be the one who decided. It ran a GRAS affirmation petition process: a company petitioned, the FDA reviewed the science, and if it agreed it published a rule affirming the substance as GRAS. Those affirmations are still on the books in 21 CFR parts 184 and 186.

The process was thorough and completely unsustainable. Every affirmation required a proposed rule and a final rule — the full federal rulemaking apparatus, for one ingredient. The backlog grew faster than the FDA could work.

So on April 17, 1997, the FDA proposed replacing affirmation petitions with something lighter: a voluntary notification program. Instead of asking permission, a company would simply tell the FDA “we have concluded this is GRAS, here is our reasoning,” and the FDA would send back a letter. That proposal was finalized on August 17, 2016.

The word doing all the work in that sentence is voluntary. From 1997 onward, the FDA operated on the honor system, and a great many companies quietly declined to participate. The practice acquired a name: self-affirmation. A manufacturer assembles its own expert panel, the panel concludes the ingredient is GRAS, the ingredient goes into the food supply, and no one outside the company ever sees the file.

Nothing about that is illegal. It is exactly what the statute permits. That is precisely why critics call it a loophole rather than a violation — you cannot enforce your way out of it, because nobody is breaking the rules. The rules are the problem.

How Big Is the Blind Spot?

The honest answer is that nobody knows, and that is the actual point of the rule.

The proposal cites an estimate that as of January 2011, more than 10,000 additives were in use in food, including roughly 1,000 substances in human food for which companies had reached their own GRAS conclusions without ever notifying the FDA. Those numbers are fifteen years old and were estimates when they were made.

The FDA's own language on this is unusually candid for a regulatory document. It writes that the agency “lack[s] a complete understanding of the number of substances currently on the market based on an independent conclusion of GRAS status,” and that this gap “prevents us from efficiently sharing knowledge and providing transparency to others, such as state regulators, food manufacturers, and consumers.”

It also names the failure mode directly: firms introduce substances “without publicly disclosing the underlying safety information,” and the FDA “sometimes becomes aware of the need to take action regarding unsafe substances in food only after adverse public health events occur.”

Read that last clause again. The current system's detection mechanism for an unsafe ingredient is people getting sick. That is the information gap the rule is trying to close, and it is a legitimate thing to want to close.

What the Proposed Rule Would Require

Stripped of the regulatory prose, the proposal does five things.

1. It flips voluntary to mandatory. New section 170.205 would require anyone introducing a purportedly GRAS substance into interstate commerce to submit a GRAS notice explaining the basis for that conclusion. The parallel provision at 570.205 covers animal feed — which matters more than it sounds, since animal feed additives can reach people through meat, milk and eggs.

2. It covers indirect additives too. The requirement reaches substances added directly to food and substances that migrate into food from packaging — the food-contact materials category. This is a meaningful scope decision; an earlier version of the regulatory agenda had appeared to drop it.

3. It makes the notices public. Submissions go into a public database, with the usual carve-out for genuine trade secrets under the Freedom of Information Act. Notably, the proposal would require a notifier to explain how a GRAS conclusion could stand even given any safety-related data being withheld from public view — a small provision with real teeth, because it makes “the reassuring data is confidential” harder to assert without scrutiny.

4. It sets review clocks. The FDA would complete an initial evaluation within 45 days to decide whether a submission is complete enough to file, then respond substantively within 180 days. The proposal adds a second 90-day extension on top of the existing one — so a full review could legitimately take up to 360 days. The FDA is candid about why: it expects the volume of notices to rise sharply, and it wants the headroom.

5. It tidies adjacent machinery. Electronic submission becomes mandatory, English translations are required for foreign-language supporting material, and the Threshold of Regulation exemption is updated.

There are sensible exceptions. You do not have to re-notify for a substance that already has an FDA “no questions” letter for that same use, or one already listed or affirmed as GRAS in parts 182, 184 or 186. Nobody has to file paperwork for vinegar.

The Streamlined Pathway for Ingredients Already on Shelves

The thousand-odd ingredients already in the food supply on a self-affirmed basis present an obvious problem: requiring a full GRAS notice for every one of them at once would bury the agency.

The proposal's answer is a time-limited streamlined submission. During a defined window after the rule takes effect, a company with an already-marketed self-affirmed ingredient can file a short form instead of a full notice: the identity of the substance, its intended uses, use levels, and evidence that it was already in interstate commerce. Not the full safety dossier — just enough to get it onto the map.

This is the pragmatic half of the rule, and it is where the promised public catalogue of what is actually in American food would come from. It is also, for the same reason, the half most likely to be attacked by industry as burdensome and by public-health advocates as too thin. Both critiques have a point: a form listing an ingredient's name and use level tells you it exists, not whether it is safe.

What the Rule Explicitly Does Not Do

Here the proposed rule is more forthright than the press release announcing it. The document states directly:

“[T]he proposed GRAS notification program generally, and proposed § 170.205(a), do not establish a premarket review program for purportedly GRAS substances… a company may continue marketing a purported GRAS substance before submitting a GRAS notice or after submitting a GRAS notice before it is filed by FDA. Similarly, a company may reach a GRAS conclusion about a new use of a substance and introduce the substance into interstate commerce before submitting a GRAS notice.”

That paragraph is the entire story of the gap between the announcement and the rule. A company can:

  1. Decide internally that a new chemical is safe;
  2. Put it in food and start selling nationwide;
  3. File the notice afterwards;
  4. Keep selling while the FDA reviews it;
  5. Keep selling even if the FDA comes back with an “insufficient basis” letter, until the FDA takes separate enforcement action.

The reason is not timidity. It is statutory. Section 201(s) of the FD&C Act says a GRAS substance is not a food additive, and section 409 — the premarket approval machinery — applies only to food additives. As long as that definition stands, the FDA has no premarket authority over anything a company sincerely believes is GRAS. Changing that requires Congress to amend the 1958 statute. A rule cannot do it.

So when Secretary Kennedy said the loophole was being closed “once and for all,” the accurate version is: the loophole is being illuminated, not closed. That is a real improvement over darkness. It is not the thing that was claimed.

“Mandatory” Without a Penalty

This is the detail we would most encourage you to sit with, because it is in the rule's own summary of its major provisions and it has been largely absent from the coverage.

What happens to a company that simply does not file? According to the proposal, the FDA “would consider such noncompliance as a factor in its prioritization of food substances for post-market review.”

That is the consequence. Not a fine. Not a recall. Not a presumption that the unnotified substance is unlawful. The company moves up a queue for review at some later date.

Peter Lurie of the Center for Science in the Public Interest, a former FDA associate commissioner, put it to STAT as: “Mandatory without enforcement is not mandatory at all.” You can disagree with his conclusion, but the underlying observation is verifiable in the text.

There is a defensible argument on the other side. A company that skips the notice is making a documented, deliberate choice to withhold information from a federal agency — and if that ingredient later turns out to cause harm, that choice becomes a very expensive exhibit in a product-liability trial. Litigation risk is a real enforcement mechanism in American food regulation, sometimes a more effective one than agency action. But it works after the harm, which is the same timing problem the rule was written to fix.

The FDA cites sections 201, 402, 409 and 701 of the FD&C Act as its authority. The load-bearing one is section 701(a), which lets the Secretary issue regulations “for the efficient enforcement” of the Act. The argument runs: to enforce the food-additive provisions, we need to know which substances are claiming the GRAS exemption; therefore requiring notice is efficient enforcement.

That argument is not frivolous. It is also not obviously winning, for three reasons that food-regulatory lawyers have already flagged.

First, the FDA has previously said the opposite. The agency has stated in the past that it lacks express statutory authority to require GRAS notices. Prior contrary positions do not bind an agency, but they are excellent material for an opposing brief.

Second, Congress knew how to require notification and chose not to here. When it created the food-contact substance notification program in 1997, Congress mandated notification explicitly. It did not do so for GRAS. Under ordinary canons of statutory construction, that silence is read as deliberate.

Third, the deference landscape has changed. With Chevron deference overruled by Loper Bright Enterprises v. Raimondo (2024), a court reviewing this rule will decide for itself what the statute means rather than deferring to the FDA's reading. That is a materially harder environment for an agency stretching general rulemaking authority into a new mandate.

None of this means the rule fails. It means litigation is close to certain, and that the rule's survival is a genuinely open question rather than a formality.

The Timeline: Roughly 2030

If you take away one practical thing, make it this. Stack the intervals the proposal itself specifies:

Add it up and the first mandatory notices land around 2030, assuming the rule is finalized roughly on schedule and survives court challenge. The public database of what is actually in the American food supply would fill in over the following years.

This is not a criticism of the FDA — it is how notice-and-comment rulemaking works, and the deliberateness is a feature when the alternative is regulation by press release. But it should recalibrate any headline suggesting your food is about to change. Nothing on any shelf changes because of this announcement, this year or next.

The Separate Ultra-Processed Food Announcement

The same day, HHS and the U.S. Department of Agriculture announced that they had submitted a proposed federal definition of “ultra-processed food” for interagency review. This is a separate action from the GRAS rule and is at an earlier stage — the text is not yet public.

A standard definition matters more than it sounds. Right now, nutrition research on ultra-processed foods leans on the NOVA classification, a research framework developed at the University of São Paulo. NOVA is useful and widely used, but it was built to categorize foods for epidemiology, not to carry regulatory weight, and its boundaries are genuinely contested. A federal definition would let school-meal standards, food-assistance rules and labeling requirements all point at the same target.

It also creates something to fight over, and the fight started immediately. Rhonda Bentz of the Consumer Brands Association, quoted by FoodNavigator, called any definition that “captures 80% of the nation's food supply while picking winners and losers… fundamentally flawed, not based on sound science and politically motivated.”

That objection deserves a fair hearing rather than dismissal. “Ultra-processed” is a real and useful concept — the epidemiology linking it to type 2 diabetes and cardiovascular disease is substantial — but it is also a category that lumps a soda together with packaged whole-grain bread. Where the line falls will decide which products get labeled, and a poorly drawn line discredits a good idea. We will cover the definition when the text becomes public.

What Supporters and Critics Are Saying

The reaction did not split cleanly along the usual lines, which is itself informative.

Acting FDA Commissioner Kyle Diamantas framed it as a visibility fix: “By proposing mandatory GRAS notifications, we are closing critical information gaps and giving the FDA greater visibility into substances entering the food supply.” Note that this framing is accurate to the rule, and notably more modest than the Secretary's.

Dariush Mozaffarian of Tufts University, a cardiologist and nutrition-policy researcher, told STAT that the policies rank among “the FDA's most consequential nutrition actions in 50 years.”

Jennifer Pomeranz of NYU's School of Global Public Health called it a significant change, making the point that post-market review is impossible for substances whose existence the FDA cannot establish.

Vani Hari, the food activist known as Food Babe and a prominent voice in the Make America Healthy Again movement, welcomed it on transparency grounds.

Among the critics, Marion Nestle, professor emerita of nutrition at NYU and one of the most durable critics of the GRAS system, gave FoodNavigator the driest assessment: “My immediate reaction was ‘that's all?’” — adding that companies “can still put what they like into the food supply until the FDA gets around to considering the notification.” She also questioned whether the FDA has the staff, given cuts to its food divisions.

Melanie Benesh of the Environmental Working Group: “True GRAS reform requires more than mandatory notification. It must also include strong, science-based safety standards and thorough, independent FDA review.”

David Kessler, FDA Commissioner under Presidents Bush and Clinton, made a different objection: that the proposals “do not tackle the root causes of chronic diseases,” because the chemicals in ultra-processed food are not the main driver of metabolic disease — the refined carbohydrates and added sugars are. This is a substantive disagreement about what actually makes people sick, and we think it is the most interesting critique of the set. It is possible to fix the additive disclosure system completely and move population health very little.

And from industry, the International Food Additives Council rejected the premise, asserting there “has never been a loophole” while saying members are ready to work with the agency. That framing is defensible as a matter of law — self-affirmation is what the statute authorizes — and beside the point as a matter of public health, which is the disagreement in miniature.

What This Means for You

Practically, in the near term: nothing. No ingredient is being removed, no label is changing, and no product is being recalled because of this announcement. Anyone telling you otherwise is ahead of the facts.

What changes is the direction of travel, and there are a few things worth knowing.

The system you have been eating under is real, not a conspiracy theory. If you have encountered the claim that companies can put chemicals in food without telling the FDA, that claim is true, it is legal, and the FDA has now written a 140-page document confirming it. That is worth knowing regardless of what happens to this rule.

Transparency and safety are different things. A public database will let researchers, journalists and state regulators find problems years earlier than “wait for an outbreak.” It will not itself remove a single questionable ingredient. Both statements are true at once.

The strategies that worked before still work. Nothing about eating whole, single-ingredient foods depends on any of this. An apple has no GRAS filing. If you want to reduce your exposure to self-affirmed additives, the reliable method is the unglamorous one — cook from ingredients rather than from packages. Our food additives guide covers the specific compounds most worth avoiding, and US vs EU food safety regulation explains why an ingredient banned in Europe can be legal in a US product from the same manufacturer — which is largely a GRAS story.

Your comment carries unusual weight right now. See below.

How to File a Public Comment

Notice-and-comment rulemaking is one of the few places in federal policy where an ordinary person's submission is legally required to be considered. The FDA must respond to substantive comments in the final rule. This is not symbolic.

Two practical notes. Comments are posted publicly and unchanged, including any personal information you put in the body — so leave out anything you would not want indexed. And form-letter campaigns count for far less than a specific one: a comment describing a concrete situation, or answering one of the questions the FDA explicitly asks, is worth more than a thousand identical signatures. The proposal specifically requests comment on alternatives, including how long the streamlined submission window for already-marketed ingredients should stay open — a narrow, genuinely undecided question where a well-argued comment could matter.

Sources and Primary Documents

The regulatory details, quoted rule text, dates, cost figures and statutory citations in this article were taken from the pre-publication text of the proposed rule itself, not from secondary coverage. Quotes from named individuals are attributed to the outlet that reported them.

  1. U.S. Food and Drug Administration. Substances Generally Recognized as Safe. Proposed rule, 21 CFR Parts 170 and 570, Docket No. FDA-2025-N-3262, RIN 0910-AJ02. Scheduled for publication in the Federal Register August 11, 2026. federalregister.gov/d/2026-16296
  2. U.S. Department of Health and Human Services. Secretary Kennedy Announces Landmark Food Policy Reforms to Advance President Trump's MAHA Agenda. August 10, 2026. hhs.gov
  3. Cueto I, Cooney E. FDA to require notice of additives in food supply for the first time. STAT News, August 10, 2026. statnews.com
  4. FDA GRAS overhaul: HHS proposes mandatory notification. FoodNavigator, August 10, 2026. foodnavigator.com
  5. FDA Proposes Mandatory GRAS Notifications for Ingredients Introduced to U.S. Food Supply. Food Safety Magazine, August 2026. food-safety.com
  6. Venable LLP. New Details Emerge on FDA's GRAS Rulemaking: Updated Timeline, Narrower Scope, and Open Questions on Statutory Authority. July 2026. venable.com
  7. Food Additives Amendment of 1958, Pub. L. 85-929, 72 Stat. 1784. Codified at sections 201(s) and 409 of the Federal Food, Drug, and Cosmetic Act.
  8. U.S. Food and Drug Administration. Substances Generally Recognized as Safe. Final rule establishing the voluntary GRAS notification program, 81 FR 54960, August 17, 2016.
  9. U.S. Food and Drug Administration. Substances Generally Recognized as Safe. Proposed rule replacing GRAS affirmation petitions with voluntary notification, 62 FR 18938, April 17, 1997.
  10. Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024).

Connections

↑ Back to Table of Contents