The Algorithm Deciding Who Gets a Chance to Live: Connecticut Must Ban AI Embryo Selection
Artificial intelligence is already ranking human embryos in Connecticut and slotting some for implantation and others for death. Is our State subsidizing it?
In July 2026, the Center for Advanced Reproductive Services—an academic affiliate of the UConn School of Medicine—became the first clinic in its national fertility network to deploy an artificial-intelligence system called Embryo Predict.
The system examines images of embryos, assigns them numerical scores, and assists embryologists in deciding which embryo to transfer.
For those who recognize that human life begins at conception, the meaning is unmistakable: an algorithm is now helping determine which human being will be given an opportunity to live and which will be left behind and even destroyed.
Connecticut must not allow this practice to become normalized. The General Assembly should prohibit the use of artificial intelligence to select embryos for implantation or to influence decisions that result in embryos being frozen indefinitely, subjected to destructive research, or discarded.
And Connecticut voters should ask every candidate for public office a direct question:
Do you support allowing artificial intelligence to help decide which embryonic human beings will be given a chance to live—and which will not?
Connecticut embryos are already being scored by AI
The Center for Advanced Reproductive Services, commonly known as CARS, operates offices in Farmington, Hartford, Branford, Middlebury, and Danbury. It is an academic affiliate of the UConn School of Medicine.
On July 14, 2026, its parent network, First Fertility, announced that CARS had begun deploying Embryo Predict, an FDA-cleared AI system developed by Alife Health.
According to the announcement, Embryo Predict analyzes microscope images of embryos developed to days 5, 6, or 7. The system was trained on thousands of embryo images and generates a “data-driven score” to assist with embryo-transfer decisions.
CARS was identified as the first clinic in the First Fertility network to launch the technology. First Fertility announcement
The announcement does not identify the particular CARS laboratory using the system. But it confirms that AI-assisted embryo selection is no longer a distant possibility for Connecticut. It is here.
First Fertility subsequently announced an expanded partnership to deploy another AI platform, called CHLOE, across multiple IVF laboratories. CHLOE combines time-lapse imaging with AI-assisted embryo and egg assessment. The announcement does not identify which Connecticut facilities have received the system, but CARS belongs to the network deploying it. First Fertility–CHLOE announcement
Other Connecticut fertility institutions are moving toward AI
Illume Fertility, which operates Connecticut offices in Norwalk, Stamford, Danbury, and Trumbull, published a detailed explanation of its approach to AI in June 2026.
Illume says it is integrating AI, robotics, and advanced genetic testing into patient care. It currently identifies an AI system used to count and measure ovarian follicles. It also says that AI may play a role in analyzing some preimplantation genetic-testing results.
Illume states that it is “actively exploring” time-lapse embryo technology and plans to introduce it in the near future. The available public evidence does not establish that Illume is presently using AI to rank embryos for transfer. But it demonstrates how quickly Connecticut fertility practices are moving in that direction. Illume Fertility’s explanation of AI in IVF
Yale is also involved in the developing field. The laboratory director for the Yale Fertility Center and Yale Medicine Greenwich Fertility identifies AI applications in IVF among his professional interests and coauthored the American Society for Reproductive Medicine’s 2026 committee opinion on AI in IVF. A Yale senior embryologist has served as vice president of ASRM’s AI Special Interest Group.
These activities do not prove that Yale currently uses AI to select embryos in patient care. They do demonstrate that AI-assisted embryo selection is becoming part of Connecticut’s fertility-medicine infrastructure.
Connecticut lawmakers should act before this technology becomes routine and largely invisible.
Human beings reduced to scores
During IVF, multiple embryos are commonly created. Each embryo is a distinct, living human organism with his or her own genetic identity.
AI systems do not see human beings. They see images, measurements, patterns, probabilities, and scores.
The embryo receiving the highest score may be transferred first. Lower-ranked embryos may remain frozen while higher-ranked embryos receive priority. Some may eventually be donated for research or discarded.
The fertility industry calls this “embryo assessment.” A more honest description is algorithmic human selection.
An embryo’s opportunity to continue developing should not depend upon whether proprietary software identifies the “right” pattern in a microscope image.
Nor is it sufficient to say that an embryologist makes the final decision. The purpose of the AI score is to influence that decision. If the algorithm affects which embryo is transferred and which is left behind, the system is participating in a decision with life-or-death consequences.
The return of eugenics in digital form
The danger grows when AI is combined with genetic testing.
Companies are developing systems that rank embryos according to predicted disease risks. Some have moved toward predictions involving intelligence, height, obesity, mental illness, and other characteristics.
In 2024, The Guardian reported that a U.S. company was offering wealthy clients embryo screening involving predicted IQ and other traits. The service reportedly cost as much as $50,000. “US startup charging couples to ‘screen embryos for IQ’”
This is eugenics repackaged as innovation.
Instead of openly declaring some human beings unfit, a company labels their predicted characteristics unfavorable. Instead of acknowledging that a child is being rejected, the industry says that an embryo received a lower score.
The language is clinical. The discrimination is real.
This presents a particular threat to people with disabilities. If embryos receive lower rankings because they may have Down syndrome, a genetic disorder, or an increased likelihood of disability, medicine is not curing a disease. It is selecting against the human being who may have that condition.
AI makes that discrimination scalable, profitable, and easier to conceal behind a supposedly objective number.
The technology is uncertain
The moral objection would remain even if AI embryo-selection systems were perfectly accurate. But they are not.
The American Society for Reproductive Medicine acknowledges that AI is entering IVF laboratories, particularly for embryo selection. Yet ASRM cautions that much of the favorable evidence is retrospective and that additional research is needed to validate its use. ASRM committee opinion
Illume Fertility acknowledges that a large 2024 study found an AI algorithm was not as accurate as experienced embryologists in selecting embryos most likely to result in pregnancy. It also acknowledges that the “hype around AI in embryo selection often runs ahead of what the evidence actually supports.”
An embryo receiving a low score from one system might receive a higher score from another. Nevertheless, that uncertain prediction may influence whether the embryo is transferred, kept frozen, or ultimately destroyed.
The industry promises precision. What it actually offers is a probability—with a human life hanging in the balance.
Connecticut should prohibit algorithmic embryo selection
Connecticut should not merely require clinics to disclose that they are using AI. Disclosure is important, but informed consent does not make an inherently unjust practice acceptable.
The General Assembly should prohibit fertility clinics, physicians, embryologists, genetic-testing companies, and their contractors from using artificial intelligence to:
- Rank human embryos for implantation or transfer;
- Assign embryos viability, desirability, or quality scores used in transfer decisions;
- Recommend that one embryo be selected over another;
- Select embryos according to predicted disabilities, diseases, intelligence, appearance, or other genetic characteristics;
- Recommend that an embryo be discarded, donated for destructive research, or denied continued preservation; or
- Make or materially influence any decision resulting in an embryo’s destruction.
The law should also prohibit clinics from sending embryo images or genetic data to outside companies for prohibited AI analysis. It should require disclosure of every automated system used in the IVF process, create enforceable recordkeeping requirements, and provide meaningful civil and licensing penalties for violations.
Nothing in such a law should prevent technology from being used to protect embryos—for example, to monitor incubator conditions, prevent identification errors, or safeguard cryogenic storage. The distinction should be clear: AI may be used to protect human life, but never to rank its worth or determine its fate.
Ask every political candidate where he or she stands
This issue should be raised during Connecticut’s 2026 elections and in every campaign that follows.
Candidates routinely speak about artificial intelligence, health-care innovation, reproductive rights, disability rights, and consumer protection. They should be required to explain how those positions apply when AI is used to rank human embryos.
Voters, journalists, advocacy organizations, and debate moderators should ask candidates:
- Do you believe artificial intelligence should be permitted to rank human embryos and influence which embryo is implanted?
- Should an AI-generated score ever be used to recommend that an embryo remain frozen, be donated for destructive research, or be discarded?
- Do you support legislation banning AI from selecting embryos according to disability, disease risk, intelligence, appearance, or other predicted characteristics?
- Will you support a bill prohibiting AI from making or materially influencing any decision that results in the destruction of a human embryo?
- Will you require Connecticut fertility clinics to disclose their use of AI, the systems they use, the criteria those systems evaluate, and what happens to lower-ranked embryos?
Candidates should not be permitted to escape behind vague statements about “following the science” or “supporting innovation.” The question is not whether AI can analyze an image. The question is whether it should be permitted to influence who is selected for continued life.
A candidate who opposes discrimination should be able to explain why discrimination against human beings at the embryonic stage is acceptable.
A candidate who claims to defend people with disabilities should be able to explain whether an algorithm should be permitted to rank embryos lower because of predicted disabilities.
A candidate who supports responsible AI should be able to identify a clearer boundary than this: computers must never be permitted to help decide which human beings are given a chance to live and which are discarded.
Connecticut must act now
Connecticut is often eager to embrace emerging technology. But innovation is not automatically progress.
A system that more efficiently determines which human beings receive an opportunity to live is not morally neutral simply because it operates through sophisticated software.
Family Institute of Connecticut believes that every human being possesses equal and inherent dignity from the moment of conception. That dignity does not depend on health, intelligence, appearance, disability status, implantation probability, parental preference, or an algorithmic score.
Every embryo is already one of us.
No child should have to satisfy a computer before being given an opportunity to live. No fertility clinic should be permitted to hide decisions affecting human lives behind a proprietary algorithm. And no artificial-intelligence system should participate in deciding which human beings will be implanted and which will be left behind for destruction.
Ask every candidate where he or she stands. Demand a clear answer. Then demand legislation prohibiting AI-assisted embryo selection and destruction in Connecticut.
