A 2024 Guardian/Hope Not Hate investigation found Heliospect Genomics used AI-driven polygenic scoring to rank IVF embryos by predicted IQ

heliospect-genomicspolygenic-embryo-selectionai-genomicseugenicsjonathan-anomalyunverified-mechanism

An undercover investigation by Hope Not Hate, reported by four named Guardian journalists, found that the US startup Heliospect Genomics used genetic data from the UK Biobank to train polygenic scores — machine-learned statistical models that combine many small genetic variants into a single predicted-trait estimate — and then applied those scores to rank IVF embryos by predicted IQ for paying clients. Heliospect's CEO, Michael Christensen, was recorded describing the pitch: "Everyone can have all the children they want and they can have children that are basically disease-free, smart, healthy[...] it's going to be great." Jonathan Anomaly, a bioethicist, is named as an in-house advisor who publicly defends the practice as "liberal eugenics."

[unverified-mechanism] — the description of how the scoring pipeline works and what claims Heliospect made about its accuracy rests entirely on one investigative piece (Tier 3: named journalists, a real venue, but several layers from primary company or scientific documentation). Sources.md's sourcing floor requires Tier 1–2 for a mechanism claim of this specificity; independent corroboration is tracked at question-corroborate-heliospect-genomics-iq-embryo-investigation.