From Stealth to Sign-Ups: Heliospect’s Embryo-IQ Pitch Goes Public

Heliospect Genomics’ embryo-screening project has moved from a secretive sales pitch into a public customer funnel. As of August 2026, the Heliospect website directs people interested in embryo screening to Herasight, and Herasight’s public service page says the company is accepting customers while promoting screening for diseases and cognitive potential.

What has not changed is the central scientific limitation: these tests estimate genetic predispositions rather than revealing a future child’s intelligence. The newly visible commercial service therefore makes the distinction between prediction and certainty more important, not less—and there is still no public evidence that a score can guarantee a particular IQ, health outcome or life trajectory.

How the private pitch became a public business

The original controversy dates to October 2024, when undercover recordings showed Heliospect presenting polygenic embryo scoring to prospective customers. The Guardian’s investigation reported that the startup was in stealth mode, had marketed analysis of 100 embryos for as much as $50,000 and claimed that selecting among ten embryos could produce an average gain exceeding six IQ points.

Those figures describe the 2023–2024 pitch, not a verified 2026 price list or outcome. The current public material does not reproduce that $50,000 offer, and the historical claim of a six-point gain should not be read as a measured result from children born after screening. No independently verified follow-up published on the customer-facing pages establishes that such a gain occurred.

The material change is availability. Herasight now invites prospective customers to book a call, asks whether they already have an IVF clinic or tested embryos, and presents genetic counseling as part of its service. Its website also promotes a predictor for cognitive ability alongside tools addressing complex diseases such as Alzheimer’s disease, diabetes and schizophrenia.

That shift matters commercially because Heliospect was previously described as still developing a service that was not publicly available. A public intake process turns an experimental proposition into an immediate purchasing decision for IVF families, even though the underlying predictions remain probabilistic and the company’s own educational notices say they are not diagnoses or guarantees.

What an embryo score can—and cannot—do

Polygenic embryo screening does not edit an embryo or insert genes associated with intelligence. It analyzes many genetic variants, combines their statistical associations into a score and compares embryos created by the same prospective parents. The result is a ranking of estimated predispositions within that family’s available group of embryos.

This is a narrower task than predicting IQ from DNA with certainty. Sibling embryos share the same parents, so the genetic differences among them are limited; the number of embryos suitable for transfer may also be small. A high-ranked embryo can fail to implant, while a transferred embryo’s eventual cognitive development will reflect education, health, family circumstances and other environmental influences that are not contained in a polygenic score.

The distinction between established testing and polygenic screening is equally important. Preimplantation testing has long been used for particular single-gene conditions and chromosomal abnormalities. Cognitive ability, by contrast, is a complex trait influenced by many genetic variants and non-genetic conditions, so its score represents an estimate across a population model rather than a diagnosis attached to one embryo.

Professional guidance has become more explicit

Since the 2024 reporting, reproductive-medicine guidance has hardened rather than validated the sales proposition. In a report released in December 2025, the American Society for Reproductive Medicine concluded that polygenic embryo screening for multifactorial diseases lacked proven clinical utility and should remain in research settings under institutional oversight.

ASRM’s assessment focused on disease risk rather than intelligence, but that boundary is itself significant. Its detailed opinion says selection for traits such as intelligence, height and eye color falls outside the scope of reproductive medicine and should not be used. It also identifies incomplete knowledge of gene–environment interactions, limited representation of diverse ancestry groups and the absence of long-term outcome evidence as unresolved problems for medical polygenic screening.

The guidance does not establish that every statistical signal is meaningless. It says the evidence is not yet sufficient to turn such signals into reliable clinical decisions. A predictor may separate groups in retrospective data while still offering uncertain value for a particular family choosing among a handful of embryos.

The purchasing decision contains several separate bets

Prospective customers are not buying a direct measurement of a child’s future intelligence. They are buying an IVF-linked analysis, a model’s ranking and counseling about how to interpret that ranking. The decision therefore depends on several uncertainties that should not be collapsed into one headline score:

• whether the predictor has been validated within families rather than only among unrelated adults;
• how accurately it performs for the parents’ genetic ancestry;
• how many viable embryos are actually available for comparison;
• whether selecting for one score changes other medically relevant risks;
• and whether any claimed benefit has been confirmed through independent, long-term outcomes.

The ancestry question is especially consequential. Models trained disproportionately on people of European ancestry can lose accuracy when applied to other populations. Even within a well-represented group, statistical association does not explain how a child’s environment will interact with inherited predispositions over decades.

The controversy now concerns a real customer pathway

The debate is no longer limited to whether a startup might eventually launch an IQ-related embryo service. Heliospect’s public route now leads prospective customers to a company that markets cognitive screening and accepts inquiries, while current professional guidance continues to reject polygenic testing as established clinical care.

That creates a sharper business and regulatory conflict. Companies can improve models and publish validation studies, but customers must make irreversible reproductive choices before decades of outcome data exist. The service may generate additional information; it cannot convert an uncertain genetic estimate into a promised child.

For IVF families, the most important question is therefore not whether intelligence has a genetic component. It is whether the specific score, tested in a relevant population and among siblings, provides enough independently demonstrated value to justify changing which embryo is transferred. Public marketing has arrived ahead of a professional consensus that the technology is ready for that role.

Also read:

• Startup Option Pools: Calculate the Pre-Money Dilution Before You Sign
• How to Read a Startup Term Sheet Before You Sign It
• Sign Paperwork to Voice Darth Vader Using AI