This fact-check began with a question from home. Two children had independently encountered videos about a new Starbucks drink. The story they brought back was vivid: customers were getting sick, and the drink contained the kind of moisture-removing gel found in packets marked DO NOT EAT.
There is enough truth in that story to make it travel. Starbucks did introduce a powder containing silica. Workers have publicly described coughs, sore throats and airborne dust. Silica gel is made from the same chemical compound named in Starbucks' statement.
Then several different risks are folded into one frightening object.
Overall verdict: Half true. Confidence: medium. The powder contains food-grade, non-crystalline silica. That does not make the finished drink a silica-gel poison, and we found no evidence that drinking it causes cancer, silicosis or a pattern of silica poisoning. The reports from baristas raise a separate occupational question that Starbucks' food-safety answer does not settle.
It is a drink format, not one mystery drink
Starbucks launched its U.S. Blended Energy Refreshers on July 14. The company named several frozen drinks, including blended passionfruit-guava, pink, mango-strawberry and mango-dragonfruit versions, and said any Refresher could be blended.
The viral object is the blending powder used to help those drinks hold a consistent texture. This matters because a claim about “the new Starbucks drink” can sound as though one cup contains a hidden contaminant. The powder is an intentional preparation ingredient used across a family of blended drinks.
The traceable viral chain begins behind the counter. A TikTok account showed workers blending five scoops of the powder into a concentrated, rubbery-looking mass. The clip used ominous music, warned viewers to stop ordering the powder and included the line “you’re gonna get cancer.” An X account then amplified the video and an alleged ingredient list. The Daily Dot's reconstruction said it could not independently verify that full list and did not endorse the health claims.
Five scoops blended as an experiment are not evidence of what happens in a standard recipe. The resulting texture may be excellent theatre and poor toxicology.
The silica is real
After the clips spread, Starbucks published a short response on July 27. It said the powder contains “a small amount” of food-grade silica, that the ingredient helps the mixture stay consistent, and that the product contains no crystalline silica.
Silica is another name used for silicon dioxide, SiO₂. U.S. food rules allow silicon dioxide as an anticaking agent when it has that technical effect, is used only as much as reasonably required and does not exceed 2 percent by weight of the food. That ceiling is a legal condition for permitted use, not a measurement of this Starbucks powder. Starbucks has not publicly stated its exact percentage.
The European Food Safety Authority reached a similar dietary conclusion by a different route. Its 2024 assessment of silicon dioxide, or E551, found no safety concern for any population group at the reported uses and use levels. The panel used a margin-of-exposure approach rather than setting an acceptable daily intake because the evidence base still had limitations, especially around nanosized aggregates.
So “approved for food” is not a magic phrase meaning every amount and every route of exposure is harmless. It does mean that the viral leap from an unfamiliar chemical name to a poisoned drink conflicts with how regulators assess this ingredient in food.
The silica-gel comparison is chemically right and practically misleading
Poison Control explains that silica gel is a form of silicon dioxide. It absorbs water, which makes it useful in the little desiccant packets shipped with medicines, electronics, clothes and other goods. The packets are labelled DO NOT EAT.
That shared chemistry is the kernel of truth. The conclusion usually attached to it is wrong.
Silicon dioxide can have different structures, particle sizes, grades and purposes. Starbucks describes a fine food-grade powder used to keep ingredients flowing and mixing. A desiccant packet contains a porous form designed to absorb moisture, commonly as beads or granules. There is no evidence that Starbucks is tipping packet beads into its blenders.
More important, DO NOT EAT is not a synonym for “poison.” Poison Control says standard clear or white silica gel is usually non-toxic if accidentally swallowed. The packet itself can be a choking hazard or, rarely, cause a digestive blockage; inhaled granules can irritate. Some coloured indicator gels have additional chemicals. Those are good reasons not to eat a packet. They are not evidence that every food containing silicon dioxide is toxic.
Sub-verdict: “It contains the same substance as moisture-removing silica gel” is mostly true, but missing the differences in form, grade, dose and use. “The packet warning proves the drink is poisonous” is false.
Drinking and breathing are different exposures
The strongest unresolved issue is behind the counter.
On July 15, Starbucks Workers United said it was hearing concerns from baristas, that the silica-based powder becomes airborne easily and that Starbucks should respond if workers were experiencing side effects. Self-described workers on Reddit and elsewhere have reported coughs, sore throats, headaches, eye irritation and chest discomfort. One customer in a Reddit discussion blamed a night of gastrointestinal illness on a blended drink.
These accounts establish that people have reported symptoms. They do not establish what caused them. A sore throat after a shift could be related to airborne blending powder, another dusty ingredient, an infection, an irritant or something else. A stomach illness after a drink has the same causation problem. No denominator tells us how many drinks were served or how many workers handled the powder without symptoms. No clinical investigation links the reports to silica.
The concern is still plausible enough to measure. The NIOSH Pocket Guide for amorphous silica lists inhalation as an exposure route, names the eyes and respiratory system as target organs and gives a recommended eight-hour exposure limit of 6 milligrams per cubic metre. It includes silica gel and precipitated amorphous silica among its related terms.
That number cannot be compared with a video. We found no public air-sampling results from Starbucks cafés, no disclosed particle-size specification, no full formula and no incident dataset. Without those, nobody outside the company can calculate a worker's dose or say whether an occupational limit was exceeded.
The cancer claim changes the kind of silica
The most alarming clips borrow hazards associated with respirable crystalline silica. That is the dust produced in activities such as cutting stone and concrete. OSHA says long-term occupational exposure can cause silicosis, lung cancer and other diseases.
But OSHA explicitly distinguishes amorphous silica, including silica gel, from crystalline silica. Starbucks says its powder contains no crystalline silica. The evidence cited for stone-cutting cancer risk cannot simply be transferred to a food-grade amorphous ingredient because both share the word “silica.”
This does not prove that a café dust cloud is comfortable or adequately controlled. Fine airborne particles can irritate even when they are not the crystalline carcinogen invoked in a viral caption. It does make the cancer and silicosis claim unsupported for the powder Starbucks has described.
What Starbucks has and has not answered
Starbucks answered the consumer ingredient question in one sentence: a small amount of food-grade silica, no crystalline silica, safe for use in food. The statement is consistent with food regulation and EFSA's dietary assessment.
It did not publish the full formulation, silica concentration, supplier specification, safety data sheet or particle-size distribution. It did not say whether it had measured airborne powder during scooping and refilling, how many symptom reports it had received, or whether handling procedures had changed. We prepared those questions for a future request for comment, but did not send it because external contact was outside the authorised scope of this fact-check draft.
We also found no public FDA recall or official illness investigation tied to the blending powder as of August 5. That is not proof that no private complaints exist. It is a reason not to turn anecdotes into an outbreak.
The useful answer for a child — and an editor
Yes, the Starbucks powder contains silicon dioxide. Yes, silica gel is a form of silicon dioxide. No, the words DO NOT EAT do not reveal a secret poison in the drink. They refer to a packet that is not food and can be a choking hazard, even though ordinary silica gel itself is usually non-toxic.
There is no solid evidence that customers are being poisoned by silica in Blended Refreshers. Claims about cancer or silicosis confuse non-crystalline food-grade silica with the crystalline dust responsible for those well-established occupational diseases.
The baristas' question deserves more than a food-label answer. If a new powder repeatedly becomes airborne and workers report irritation, the decisive evidence is mundane: the product specification, the handling procedure and air measurements taken during a normal shift.
The viral videos found a real ingredient. They aimed the warning at the cup before anyone published the measurements behind the counter.