In the still, often emotionally charged space where a family begins, hope meets uncertainty. For intended parents standing before a chart of embryos, each representing a world of potential, the question has long remained: Which one will give us the best chance?

That decision is never purely clinical. In my practice, I’ve talked with countless clients who carry not just the burden of biology, but the weight of years: failed cycles, financial strain, and a longing that runs deep. For many, selecting an embryo is the most consequential decision they will ever make.

And now, that decision is increasingly shaped by machines.

Recently, Time magazine reported a pregnancy achieved with the aid of STAR, an AI-powered system developed by Columbia University researchers to identify viable sperm in men with azoospermia. While STAR currently targets only a specific form of male infertility, it represents a broader trend: artificial intelligence is entering the reproductive space not only to diagnose, but to guide decision-making at some of its most delicate junctures.

In parallel, tools like CHLOE, developed by the Israeli medtech firm Fairtility (not to be confused with Future Fertility’s VIOLET system), offer algorithmic rankings of embryos based on time-lapse video analysis. These technologies are not yet ubiquitous, nor are they uniformly regulated. But their implications for law, ethics, and patient care are already profound.

A New Player in the Lab: CHLOE and the Rise of Embryo-Ranking AI

CHLOE monitors embryo development via time-lapse imaging, analyzing parameters such as cell division timing, morphology, and symmetry to predict implantation potential. It generates a ranked list intended to assist embryologists, not override them.

To many, this might seem like a helpful refinement: a second set of (digital) eyes. But legally and ethically, this marks a shift. Consider this hypothetical scenario: A clinic has three viable embryos. An embryologist recommends one based on traditional morphological assessment; CHLOE ranks another highest based on its algorithmic analysis. Which voice prevails? And if the outcome isn’t what was hoped for, who bears responsibility?

The moment an algorithm begins shaping embryo selection, we move beyond the realm of professional judgment into a domain where proprietary models, trained on unpublished datasets and shaped by clinic-specific variables, begin to influence life’s most intimate beginnings. This technical evolution requires more than trust in innovation. It demands ethical accountability and fully informed decision-making at every step.

Regulatory Oversight: A Shifting Landscape

In the United States, oversight by the Food and Drug Administration (FDA) over AI tools in reproductive medicine is evolving. The FDA regulates AI tools in ART when they meet the criteria for Software as a Medical Device (SaMD), particularly if the tool claims to diagnose, cure, mitigate, or prevent disease. However, the regulatory classification of specific embryo-ranking AI systems varies significantly. As of early 2025, there is no publicly available record indicating that CHLOE has received FDA clearance, and it likely operates under clinical decision support parameters rather than making diagnostic claims.

The global regulatory picture is similarly fragmented. iDAScore, developed by Vitrolife, has received CE mark certification in Europe. Others remain in pre-clinical phases, reflecting the patchwork nature of international AI oversight.

Without comprehensive regulatory frameworks or transparent disclosure standards addressing AI use in embryo selection, accountability remains diffuse, and the burden of risk often falls on individual clinics and patients. While medical societies like ASRM have hosted discussions about AI applications in reproductive medicine, specific binding guidance on disclosure or consent remains limited as of early 2025.

This regulatory uncertainty creates potential risks:

• Lack of standardized consent language specific to AI use in embryo selection
• Unclear communication of the risks and limitations of algorithmic involvement
• Ambiguity around what constitutes “medical advice” versus “decision support”
• Limited legal recourse if outcomes fall short of expectations

ART clients must be aware of and prepare for these regulatory gaps, not when they are already navigating a dispute, but before treatment begins. By that point, choices about consent forms, embryo disposition, or data use may already be in place, often without legal review. Collectively, ART professionals should anticipate and disclose where ambiguity may arise and help clients make informed, strategic decisions from the outset.

Cross-Border Complications: Fertility Without Borders, Law Without Uniformity

With the rise of AI tools like CHLOE, cross-border fertility journeys now intersect with a loosely regulated layer of decision-making with consequences which must be considered.

Consider the emerging and largely untested regulatory landscape:

• Germany: The Embryo Protection Act prohibits embryo selection unless justified by clear medical indication. The law has evolved since its 1990 enactment. For instance, it now allows preimplantation genetic diagnosis in specific cases as of 2011. However, AI-assisted embryo selection may fall outside permissible use unless tightly justified. The application of these restrictions remains legally untested, but the baseline suggests potential compliance challenges.
• Canada: The Assisted Human Reproduction Act prohibits sex selection for non-medical purposes and establishes criminal prohibitions for certain activities. While the Act does not broadly restrict embryo selection to medical purposes only, Canadian clinics often self-regulate based on ethical guidelines. The intersection of AI-based embryo ranking with Canada’s consent requirements and prohibited activities remains largely unaddressed by regulators.
• Europe (GDPR): Embryo imaging data, depending on associated metadata, may qualify as health data under the General Data Protection Regulation (GDPR). Of particular concern are GDPR’s automated decision-making provisions (Articles 22 and 13 to 15), which may apply when AI is used without meaningful human oversight. These provisions require clarification in the ART with AI context, particularly with regard to patient rights concerning algorithmic decisions. Specific enforcement in ART contexts, however, remains unclear.

Now imagine a plausible but legally untested scenario: A French intended parent uses CHLOE-assisted embryo selection at a U.S. clinic. The remaining embryos are later shipped to a clinic in Spain.

• Could the U.S.-based AI selection conflict with French or Spanish legal frameworks?
• Might transport of algorithmically selected embryos raise compliance issues under local rules?
• If a dispute arises, which jurisdiction governs the AI-assisted treatment decisions?

These questions are not hypothetical. They represent the next wave of cross-border legal complexities, and they underscore why global clients need counsel fluent in both emerging technology and international law.

Note: The above scenarios represent potential legal considerations based on current frameworks. Actual enforcement positions may vary and should be verified with local counsel.

Learning from Precedent: What AI in Medicine Can Teach Fertility Law

Legal precedent addressing AI in medical decision-making is emerging in adjacent fields, though direct precedent in ART remains limited. In medical malpractice litigation involving AI-assisted diagnostics in radiology, oncology, and dermatology, courts have begun to examine:

• Whether physicians appropriately relied on AI input
• Whether patients were informed of AI involvement in care
• Whether AI tools were responsibly described and understood

In fertility law, the same questions loom:

• If an embryo ranked highest by AI fails to implant, could over-reliance on the algorithm be seen as malpractice?
• If clinics use AI tools without obtaining clear, informed consent, could liability attach?
• If clinicians override AI recommendations, what standard of care governs that judgment?

While formal litigation specifically around AI-assisted embryo selection has not yet surfaced in publicly reported cases as of early 2025, that does not signal safety. It signals latency. These are emerging legal tensions that ART law must prepare for.

The Lawyer’s Role in the Age of Algorithmic Fertility

Innovation needs legal architecture. And in ART, that architecture must account not just for clinical science but for the families science is meant to serve.

Our clients don’t speak in algorithms. They speak in dreams: of passports, birth records with tiny footprints, bedtime stories, and hospital bracelets. That’s why our role isn’t just transactional. It’s deeply human.

At IFLG, we help clients:

• Understand the legal implications of emerging reproductive technologies
• Navigate international ART frameworks with confidence
• Ensure informed consent reflects both scientific and ethical complexity
• Minimize legal risk for agencies and clinics as standards evolve

The Future We’re Building—With Families at the Center

The rise of STAR and CHLOE isn’t just about innovation. It’s about the responsibility that follows, and the legal and ethical scaffolding we must build to keep pace with progress.

These technologies represent meaningful scientific advances. STAR offers new hope for men with azoospermia. CHLOE equips embryologists with analytical tools to support, not replace, their expertise. Taken together, they reflect a new convergence of machine learning and family building. And with each technical leap forward, law will follow.

At IFLG, we will continue to guide our clients with the same combination of precision and empathy that has defined our practice for over thirty years.  The heart of every legal question in ART isn’t just a statute or contract; it’s a child, a story, and a future that deserves our most careful stewardship.