HHS removed FDA’s “not dangerous” cellphone radiation pages because animal toxicology, WHO-commissioned reviews, and federal court findings make blanket safety assurances scientifically indefensible.
Why RFK Jr. Was Correct to Remove Blanket Assurances About Cell Phone Radiation Safety
In January 2026, the U.S. Department of Health and Human Services (HHS), led by Secretary Robert F. Kennedy Jr., announced a new government study on cellphone radiation—while the Food and Drug Administration (FDA) removed older webpages that had communicated broad assurances that cellphones were “not dangerous.”
Critics framed the move as anti-science. A more accurate interpretation is the opposite: removing blanket assurances is a science-based correction to public health messaging. When the evidence base contains credible hazard signals—especially from well-conducted animal carcinogenicity studies—government agencies should not present certainty they do not actually have.
This is the core point: precautionary public health standards are built to protect populations, not to defend the strongest possible “all-clear” statement. And in regulatory science, animal toxicology is not a side note—it is often the foundation.
1) Animal toxicology is how precautionary standards are built
If you want standards that protect real people—including children, pregnant people, heavy users, and medically vulnerable groups—you do not wait for perfect, decades-long human experiments (which are often impossible). You use a weight-of-evidence framework in which:
-
Animal bioassays identify hazard potential under controlled conditions.
-
Mechanistic studies test biological plausibility.
-
Human epidemiology helps estimate risk in real-world settings, while acknowledging it can miss signals due to exposure misclassification, latency, and rapidly changing technologies.
A WHO-commissioned systematic review of animal cancer studies (Mevissen et al., Environment International, 2025) states the principle plainly: “Cancer bioassays conducted in experimental animals are commonly used to identify potential human carcinogens.”
That is not activism. That is the standard logic of protective regulation.
So if animal carcinogenicity evidence is strong enough, the scientifically conservative position is not “safe.” The conservative position is: “we cannot provide blanket assurance.”
2) The animal evidence is no longer a single outlier—it is convergent
A. The U.S. National Toxicology Program (NTP) found tumor signals
NTP’s cellphone radiofrequency radiation (RFR) program reported that high exposure was associated with:
-
Clear evidence of malignant schwannomas in the hearts of male rats
-
Some evidence of malignant gliomas in the brains of male rats
NTP also published follow-up work reporting increased DNA damage in specific tissues in exposed animals (with the important scientific note that DNA damage is a plausible upstream contributor to carcinogenesis if not repaired).
NIH’s public summary of the final NTP reports likewise emphasized “clear evidence” of cancerous heart tumors in male rats under high exposures, while also noting that exposure conditions in the studies differ from typical human use (whole-body exposure, higher dose/duration).
B. A WHO-funded animal cancer review judged “high certainty” for two tumor types
The WHO-commissioned animal cancer review (Mevissen et al., 2025) went further than simply “mixed findings.” In its evidence assessment:
-
The certainty of evidence (CoE) for an increased risk of glioma was judged high.
-
The certainty of evidence for malignant heart schwannomas in male rats was judged high, based on two positive bioassays.
Two details here matter for public messaging:
-
“High certainty” animal carcinogenicity is incompatible with blanket consumer assurance.
-
The review explicitly highlights unresolved scientific complexities for extrapolation, including whether a monotonic dose-response is even the correct assumption for RF-EMF carcinogenic effects.
That is directly relevant to your point about nonlinear exposure-response behavior. Even a mainstream systematic review is acknowledging that simplistic “dose equals heat equals harm” models may not capture what matters.
C. The Ramazzini study adds independent support for similar tumor types
The Ramazzini Institute’s long-term rat study (Falcioni et al., 2018) reported increased incidence of heart schwannomas and brain glial tumor findings in rats exposed to GSM-modulated RFR.
Whether one agrees with every methodological choice, the important public health fact is convergence on the same tumor class (schwannomas) across major animal research efforts.
Bottom line: Once a hazard signal is replicated or echoed across large animal programs and then graded at high certainty in a WHO-commissioned evidence review, a government webpage implying categorical safety is no longer scientifically defensible.
3) “Non-thermal effects” are exactly why blanket assurances are reckless
A key reason the public debate stays stuck is that “cell phone radiation” is often framed as if the only scientifically plausible harm pathway is thermal heating.
But regulatory and judicial records show that the non-thermal question has been explicitly recognized as unresolved and contested—not settled.
In Environmental Health Trust v. FCC (D.C. Circuit, 2021), the court explained:
-
RF exposure limits are designed to protect against thermal effects, not non-thermal effects.
-
At lower levels, RF radiation “might” cause other biological effects and whether such non-thermal effects are harmful is a subject of debate.
This matters because many “blanket assurance” statements effectively tell the public: “if it doesn’t heat you, it can’t affect you.” That is an assumption, not a demonstrated universal truth.
Mechanistic literature proposes pathways consistent with sub-thermal biological interaction, including oxidative stress and downstream effects on signaling and DNA integrity (not presented here as proven for humans, but presented as biologically plausible and under active study).
So even if you believe the dominant risk is low, the honest scientific posture is not certainty—it is uncertainty with protective guidance.
4) The legal record shows “blanket assurance” language became a regulatory crutch
One of the strongest evidence-based arguments for removing the FDA’s categorical language is not rhetorical—it is documented.
The FDA “Cell Phones” page still contains statements like:
-
“The weight of scientific evidence has not linked cell phone radio frequency radiation with any health problems.”
-
“The scientific evidence does not show a danger … including children and teenagers.”
In EHT v. FCC, the D.C. Circuit described how the FCC leaned on these kinds of FDA statements when refusing to update policy—and the court criticized that posture. The court noted that the FCC cited an FDA webpage with the “weight of scientific evidence has not linked…” statement, and relied on the “no danger…including children and teenagers” line, calling it “conclusory and unexplained” for purposes of reasoned decision-making under the APA.
This is the governance problem:
-
If FDA/HHS posts sweeping assurances, those statements do not just inform consumers.
-
They become institutional cover for leaving outdated assumptions and testing regimes in place—especially when standards are framed around thermal effects.
Removing blanket assurances is a way of preventing public-facing language from being used as a substitute for real, updated scientific reasoning.
5) Human data: “often null” does not equal “assured safe,” and heavy-use signals still matter
It is true that a WHO-commissioned systematic review of human observational evidence (Karipidis et al., 2024) reported meta-estimates around null for several tumor types (e.g., glioma mRR ≈ 1.01), and concluded no association in the included data.
But two science-based points are routinely lost when this becomes “case closed” messaging:
A. Even well-done human reviews cannot prove a universal negative
A null meta-estimate does not logically support the claim that there is no risk for any subgroup, under any usage pattern, over decades, across evolving exposure conditions.
In public health, “safe” is not proven by “we didn’t detect a clear increase.”
B. High-use tails show signals—and “heavy use” definitions are dated
INTERPHONE’s published data reported that, in the highest decile of cumulative call time (≥ 1,640 hours), the odds ratio for glioma was 1.40 (95% CI 1.03–1.89).
This does not prove causality (the paper discusses bias/implausible reporting issues). But it does mean the literature contains non-trivial signals at the high end—exactly where precautionary policy should focus.
6) Why RFK Jr.’s move is evidence-based—even if you disagree with his politics
If the question is: Was it scientifically appropriate to remove blanket “cellphones are not dangerous” style assurances while launching a new study? the answer is yes, for a simple reason:
Blanket assurance is a stronger claim than the evidence supports
In January 2026, Reuters reported that HHS announced a new study and that FDA removed old webpages stating cellphones are not dangerous.
That sequence is rational and aligned with scientific integrity:
-
If you are saying “we need updated research,” you should not simultaneously tell the public the question is settled.
-
If animal bioassays show tumor signals and a WHO-commissioned animal review grades two endpoints at high certainty, you should not present categorical safety language.
-
If a federal appeals court has already criticized reliance on conclusory safety assertions to dismiss complex evidence, you should tighten and modernize the public-facing language.
The most accurate statement is not “phones are unsafe,” but “there is no absolute assurance of safety”
No honest scientific institution can provide a guarantee that a ubiquitous exposure is “safe” in all conditions—especially when:
-
animal carcinogenicity evidence exists at high certainty for specific tumor types,
-
mechanisms have been proposed and studied that do not require heating as the only interaction pathway,
-
and dose-response behavior and exposure metrics remain scientifically contested.
That is the scientific “truth” in policy terms: blanket assurance is not available.
7) What responsible public guidance should say instead (and why this protects Americans)
A science-aligned, precautionary messaging model looks like this:
-
What we know: Most human studies do not show a clear increase in brain tumor risk at the population level, but there are limitations and uncertainty—especially at the highest exposure patterns.
-
What animal evidence shows: Large animal studies and a WHO-commissioned review report carcinogenicity signals with high certainty for specific endpoints.
-
What that means for safety statements: It is not scientifically valid to issue blanket assurances such as “not dangerous,” “no health problems,” or “no danger including children.”
-
What people can do now (optional precaution): Use speakerphone, wired headsets, text more, avoid carrying an active phone against the body, and reduce unnecessary exposure for children.
Precaution is not panic. It is how public health protects people when the downside of modest exposure reduction is low and the uncertainty is non-trivial.
FAQ
Did RFK Jr. ban cell phones or declare them proven dangerous?
No. The reported action was removal of older FDA webpages communicating broad assurances, alongside an HHS plan to study health impacts—i.e., a shift in public messaging and research posture.
Why do animal studies matter more than people think?
Because controlled animal cancer bioassays are a standard tool for identifying potential human carcinogens—and are explicitly recognized as such in WHO-commissioned evidence reviews.
What did the NTP cell phone radiation study find?
NTP reported clear evidence of malignant heart schwannomas in male rats and some evidence of malignant gliomas in male rats under high RFR exposures; it also reported follow-up findings consistent with DNA damage in some tissues.
What did the WHO-funded 2025 animal cancer review conclude?
It judged the certainty of evidence as high for increased risk of glioma and high for malignant heart schwannomas in male rats, while also noting major complexities in extrapolating risk to humans.
Do WHO-commissioned human studies say there’s no risk?
A WHO-commissioned human evidence review reported no association in the included observational evidence overall; that is not the same as proving universal safety for all exposure patterns.
Why can’t agencies promise “cell phones are safe”?
Because “safe” is an absolute claim. With high-certainty animal carcinogenicity endpoints, unresolved mechanism questions (including non-thermal debates), and evolving real-world exposure patterns, no agency can responsibly provide a blanket guarantee.
