Infertility has become a major public health concern affecting millions of couples worldwide. The World Health Organization (WHO) estimates that approximately one in six adults experiences infertility during their reproductive years, making access to effective fertility care an important component of reproductive health services (World Health Organization [WHO], 2023). Factors such as delayed childbearing, lifestyle changes, environmental pollution, obesity, chronic diseases, and genetic disorders have contributed to the growing demand for fertility treatment across the globe.
Assisted reproductive technology (ART), especially in vitro fertilization (IVF), has transformed infertility management by enabling many individuals and couples to achieve pregnancy despite complex reproductive challenges. Since the birth of the first IVF baby in 1978, advances in reproductive medicine have significantly improved fertilization techniques, embryo culture, cryopreservation, and laboratory procedures. Nevertheless, IVF success rates remain variable, and many patients require several treatment cycles before achieving pregnancy. Clinical decision-making often depends on the experience of embryologists and fertility specialists, which introduces the possibility of subjective interpretation and variability in treatment outcomes (Zaninovic & Rosenwaks, 2020).
Against this background, artificial intelligence has recently emerged as a promising tool for addressing these limitations. AI refers to computer systems capable of performing tasks that normally require human intelligence, including learning from data, recognizing patterns, making predictions, and supporting complex decision-making. In reproductive medicine, AI applications include embryo image analysis, sperm quality assessment, prediction of IVF success, individualized ovarian stimulation protocols, automation of laboratory procedures, and analysis of large clinical datasets (Esteva et al., 2019; Meskó & Görög, 2020; VerMilyea et al., 2020).
The increasing availability of digital imaging, electronic medical records, and machine learning algorithms has accelerated the integration of AI into fertility clinics. Several studies have reported that AI-assisted embryo selection can improve the prediction of embryo viability and reduce variability between embryologists. AI has also shown potential in optimizing treatment protocols by identifying patient-specific factors associated with successful pregnancy outcomes (Curchoe & Bormann, 2019).
Despite these promising developments, the use of AI in ART raises important ethical and legal questions. Concerns have been raised regarding algorithm transparency, data privacy, informed consent, bias in training datasets, accountability for AI-assisted clinical decisions, and unequal access to advanced fertility technologies. In addition, regulatory frameworks governing AI in healthcare differ considerably across countries, creating challenges for international standards and supervision (European Union, 2024).
Given the rapid expansion of AI applications in reproductive medicine, there is a need to evaluate the current evidence critically while examining the ethical and regulatory implications of these technologies. Existing reviews often focus primarily on clinical performance or technical aspects, with limited integration of legal and ethical considerations. A comprehensive assessment is therefore necessary to support clinicians, policymakers, researchers, and regulatory authorities in developing responsible approaches to AI adoption within fertility care.
This review examines the current clinical applications of AI in assisted reproductive technology, evaluates the available evidence regarding its effectiveness, explores ethical and legal challenges, reviews existing regulatory frameworks, and discusses future directions for the safe and equitable integration of AI into reproductive medicine.
This review adopted a narrative literature review approach to examine the current role of artificial intelligence (AI) in assisted reproductive technology (ART). A narrative review was considered appropriate because it allows the integration of findings from clinical, ethical, legal, and regulatory studies to provide a comprehensive understanding of the subject.
Relevant literature was identified through electronic databases, including PubMed, Scopus, Web of Science, Google Scholar, ScienceDirect, and SpringerLink. Additional information was obtained from reports published by the World Health Organization (WHO), the European Commission, the U.S. Food and Drug Administration (FDA), the Human Fertilisation and Embryology Authority (HFEA) in the United Kingdom, and the American Society for Reproductive Medicine (ASRM).
The search focused primarily on publications published between 2020 and 2026 to ensure that the review reflected the most recent developments in AI and reproductive medicine. A limited number of earlier landmark studies were also included where they provided important background information on IVF, machine learning, or AI applications in healthcare.
Keywords used during the literature search included artificial intelligence, assisted reproductive technology, in vitro fertilization, machine learning, deep learning, embryo selection, fertility treatment, reproductive medicine, AI ethics, data privacy, clinical decision support, and AI regulation. Different combinations of these keywords were used to retrieve relevant articles.
Studies were included if they:
1. discussed AI applications in ART or IVF;
2. reported clinical, ethical, legal, or regulatory aspects of AI in reproductive medicine;
3. were published in peer-reviewed journals or by recognized international organizations; and
4. were written in English.
Publications that focused solely on technical algorithm development without discussing clinical relevance, duplicate studies, conference abstracts with limited data, and non-English articles were excluded.
The selected literature was critically reviewed and organized into key themes, including the evolution of AI in reproductive medicine, clinical applications, benefits, ethical concerns, legal and regulatory frameworks, current limitations, and future directions. Rather than simply summarizing previous studies, this review compares existing evidence, identifies areas of agreement and disagreement, and highlights important knowledge gaps that require further research.
The methodology thus aimed to provide a balanced and up-to-date synthesis of current evidence to support healthcare professionals, researchers, policymakers, and regulators involved in the development and implementation of AI-assisted fertility care.
Assisted reproductive technology (ART) refers to medical procedures used to help individuals or couples achieve pregnancy when natural conception is difficult or impossible. According to the World Health Organization, infertility affects approximately 17.5% of adults worldwide, demonstrating the growing need for accessible and effective fertility services (WHO, 2023). ART has therefore become an important component of modern reproductive healthcare.
The most common form of ART is in vitro fertilization (IVF), in which mature eggs are retrieved from the ovaries and fertilized with sperm in a laboratory. After fertilization, one or more embryos are transferred into the uterus with the aim of establishing a successful pregnancy. Other ART procedures include intracytoplasmic sperm injection (ICSI), embryo cryopreservation, donor egg or sperm treatment, preimplantation genetic testing (PGT), and frozen embryo transfer.
Over the past four decades, advances in laboratory techniques have significantly improved ART results. Innovations such as time-lapse embryo imaging, improved culture media, vitrification (rapid embryo freezing), and genetic screening have improved pregnancy and live birth rates while reducing treatment-related risks. Despite these improvements, success rates remain influenced by several factors, including maternal age, embryo quality, ovarian reserve, sperm quality, underlying medical conditions, and the experience of fertility specialists.
One of the most challenging aspects of IVF is selecting the embryo with the highest potential for implantation and healthy development. Traditionally, embryologists assess embryo quality by examining characteristics such as cell number, symmetry, fragmentation, and developmental stage under a microscope. Although these methods are widely accepted, they rely on human observation and professional judgment, which can lead to variability between laboratories and clinicians (Curchoe & Bormann, 2019; Gardner et al., 2015).
Similarly, determining the most appropriate ovarian stimulation protocol, predicting IVF success, and assessing sperm quality often require the interpretation of multiple clinical variables. These complex decisions create opportunities for human error and inconsistency, especially when large amounts of patient data must be analyzed within a limited time.
Consequently, the rising use of digital imaging systems, electronic health records, and advanced data analytics has created an environment in which AI can support reproductive specialists by analyzing complex datasets more efficiently than traditional methods. Instead of replacing clinicians, AI is increasingly viewed as a decision-support tool that enhances diagnostic accuracy, standardizes laboratory assessments, and promotes more personalized fertility treatment.
As AI technologies continue to evolve, they are expected to play an important role throughout the entire ART process, from patient assessment and treatment planning to embryo selection and the prediction of pregnancy outcomes. The following section discusses how AI has evolved from a research concept into a practical tool in modern reproductive medicine.
Artificial intelligence (AI) has developed rapidly over the years and is now becoming an important part of reproductive medicine. Although AI has been used in healthcare for many years, its application in assisted reproductive technology (ART) has gained significant attention because of advances in computing power, digital imaging, and machine learning. These developments have enabled fertility clinics to analyze large amounts of clinical and laboratory data more accurately and efficiently than traditional methods allow.
The earliest applications of AI in ART focused mainly on predicting the likelihood of successful pregnancy using patient characteristics such as age, body mass index, hormone levels, ovarian reserve, and previous IVF outcomes. These prediction models helped clinicians estimate treatment success but were limited by small datasets and relatively simple statistical methods (VerMilyea et al., 2020).
Recent advances in machine learning (ML) and deep learning (DL) have significantly expanded the role of AI in fertility care. Machine learning enables computer systems to learn from existing data and improve their performance without being explicitly programmed for every task. Deep learning, a more advanced form of machine learning, uses artificial neural networks to recognize complex patterns in medical images and clinical records. These technologies have proven useful in analyzing embryo images, sperm characteristics, ultrasound scans, and patient health records (Curchoe & Bormann, 2019).
One of the major breakthroughs in AI-assisted reproduction has been the development of time-lapse embryo imaging systems. These systems continuously capture images of developing embryos without disturbing laboratory conditions. AI algorithms analyze thousands of images to identify developmental patterns associated with successful implantation and live birth. Compared with conventional microscopic assessment, AI-assisted image analysis provides more objective and consistent evaluations while reducing differences between embryologists (Khosravi et al., 2019; Tran et al., 2019).
AI has also transformed sperm analysis. Traditionally, sperm assessment relied on the manual examination of concentration, motility, and morphology under a microscope. AI-powered computer-assisted semen analysis (CASA) systems can rapidly evaluate thousands of sperm cells, improving accuracy, reducing observer variation, and shortening laboratory processing time (Dimitriadis et al., 2022).
Another important development is the use of AI for personalized fertility treatment. Modern AI systems can combine clinical history, hormone profiles, ultrasound findings, genetic information, and previous treatment outcomes to recommend individualized ovarian stimulation protocols and predict patient responses. This personalized approach may improve treatment effectiveness while reducing complications such as ovarian hyperstimulation syndrome (OHSS).
Cloud computing and electronic health records have further accelerated AI adoption by allowing fertility clinics to collect and analyze larger datasets. With access to data from several centers, AI models can continuously improve their predictive performance and support evidence-based clinical decision-making. However, the quality of AI predictions depends heavily on the quality, diversity, and completeness of the data used to train these systems. Poor-quality or biased datasets may reduce accuracy and limit their usefulness in different populations (High-Level Expert Group on Artificial Intelligence [AI HLEG], 2019).
Despite these advances, AI remains a clinical decision-support tool rather than a replacement for fertility specialists. Human expertise is still essential for interpreting AI-generated recommendations, communicating treatment options to patients, and making final clinical decisions. Professional organizations, including the American Society for Reproductive Medicine (ASRM) and the Human Fertilisation and Embryology Authority (HFEA), emphasize that AI should complement professional judgment rather than replace it (American Society for Reproductive Medicine [ASRM], 2026).
The evolution of AI in ART thus reflects a gradual shift from experimental research to routine clinical support. As algorithms continue to improve and larger datasets become available, AI is expected to play an even greater role in enhancing the accuracy, efficiency, and personalization of fertility treatment.
Artificial intelligence has been integrated into various stages of assisted reproductive technology (ART). Rather than replacing fertility specialists, AI supports clinical decision-making by analyzing large volumes of medical and laboratory data quickly and consistently. The most important clinical applications are discussed below.
Embryo selection is one of the most widely studied applications of AI in reproductive medicine. Choosing the embryo with the highest implantation potential is essential for improving pregnancy and live birth rates while reducing the need to transfer multiple embryos.
Traditionally, embryologists evaluate embryo quality by examining characteristics such as cell number, symmetry, fragmentation, and developmental stage under a microscope. Although this method has been used successfully for many years, it is subjective and may produce different results among embryologists. AI addresses this limitation by analyzing thousands of embryo images using machine learning and deep learning algorithms to identify subtle developmental features that may not be easily recognized by the human eye (Curchoe & Bormann, 2019).
Time-lapse imaging systems further improve this process by continuously monitoring embryo development without removing embryos from the incubator. AI algorithms assess developmental patterns and rank embryos according to their likelihood of implantation and successful pregnancy. Several studies have reported that AI-assisted embryo assessment can improve consistency and reduce observer variability, although further multicenter validation is still required before widespread adoption (Salih et al., 2023).
Predicting the likelihood of IVF success is another important application of AI. Fertility treatment outcomes are influenced by multiple factors, including maternal age, body mass index, ovarian reserve, hormone levels, sperm quality, previous pregnancy history, and lifestyle factors. Analyzing these variables simultaneously can be challenging using traditional statistical methods.
AI models can integrate these complex datasets to estimate the probability of fertilization, embryo implantation, clinical pregnancy, and live birth. Such predictions help clinicians provide realistic counseling to patients and develop individualized treatment plans. AI-based prediction tools may also reduce unnecessary treatment cycles by identifying patients who may benefit from alternative interventions before undergoing IVF.
Although prediction models continue to improve, their accuracy depends on the quality and diversity of the data used during algorithm development. Models trained using data from a single fertility center may perform less effectively when applied to different patient populations.
Male infertility contributes to nearly half of infertility cases worldwide. Conventional semen analysis involves assessing sperm concentration, motility, morphology, and viability under a microscope. This process can be time-consuming and may be affected by differences in laboratory personnel and assessment techniques.
AI-powered computer-assisted semen analysis systems automatically evaluate sperm characteristics with greater speed and consistency. Deep learning algorithms can distinguish normal from abnormal sperm cells, detect subtle morphological changes, and improve the accuracy of sperm selection for intracytoplasmic sperm injection (ICSI). These technologies may enhance laboratory efficiency while reducing observer-related errors (Dimitriadis et al., 2022).
Researchers are also exploring AI applications for predicting male fertility on the basis of genetic markers, hormonal profiles, and semen quality, although these approaches remain largely investigational.
Controlled ovarian stimulation is a critical stage of IVF treatment. Determining the optimal medication type and dosage is essential for obtaining an adequate number of mature eggs while minimizing complications such as ovarian hyperstimulation syndrome (OHSS).
AI systems analyze patient-specific information, including age, hormone levels, ovarian reserve tests, ultrasound findings, and previous treatment responses, to recommend individualized stimulation protocols. Personalized treatment may improve egg retrieval outcomes, reduce medication-related complications, and increase the efficiency of fertility treatment (European Society of Human Reproduction and Embryology [ESHRE] Guideline Group on Ovarian Stimulation, 2026).
Despite encouraging findings, clinicians continue to rely on their experience when making final treatment decisions, because AI recommendations require careful interpretation within the context of each patient’s clinical condition.
AI is increasingly used to improve quality assurance within IVF laboratories. Automated systems can monitor incubator conditions, detect equipment abnormalities, track laboratory procedures, and analyze large datasets to identify factors associated with successful treatment outcomes.
These technologies reduce manual workload, minimize procedural errors, and support standardized laboratory practices across fertility centers. Automated documentation also improves traceability, an important aspect of patient safety and regulatory compliance (European Union, 2024).
Laboratory automation should, however, be supported by regular system validation, cybersecurity safeguards, and continuous human supervision to ensure reliable performance.
Artificial intelligence is also supporting advances in reproductive genetics. AI algorithms can assist in analyzing results from preimplantation genetic testing (PGT), helping clinicians identify embryos with chromosomal abnormalities more efficiently. AI may also contribute to identifying genetic patterns associated with recurrent implantation failure, recurrent pregnancy loss, and inherited reproductive disorders.
As genomic technologies continue to develop, AI is expected to play a larger role in precision reproductive medicine by integrating genetic, clinical, and laboratory information to support personalized treatment decisions. Nevertheless, the interpretation of genetic findings requires careful ethical consideration and professional genetic counseling to avoid inappropriate clinical decisions.
Current evidence therefore suggests that AI has the potential to improve several aspects of assisted reproductive technology, including embryo selection, IVF outcome prediction, sperm analysis, ovarian stimulation, laboratory management, and genetic assessment. These applications may enhance treatment efficiency, reduce human error, and support personalized fertility care.
AI is not without limitations, however. Many existing models require further clinical validation, and their performance may vary across different healthcare settings and patient populations. Consequently, AI should be regarded as a valuable decision-support tool that complements, rather than replaces, the expertise of fertility specialists.
Artificial intelligence has demonstrated considerable potential to improve the quality and efficiency of assisted reproductive technology (ART). Although AI is still evolving, current evidence suggests that it can support clinical decision-making, improve laboratory performance, and contribute to more personalized fertility treatment. The available evidence also indicates, however, that AI should be implemented cautiously, because further validation is required before it becomes a universal standard of care.
One of the greatest advantages of AI is its ability to analyze large and complex datasets within a short period. Fertility specialists often consider numerous factors when planning treatment, including age, hormone levels, ovarian reserve, embryo quality, medical history, and previous IVF outcomes. AI can integrate these variables simultaneously and identify patterns that may not be immediately apparent through conventional analysis.
Several studies have shown that AI-assisted prediction models can estimate implantation and pregnancy outcomes with encouraging accuracy. These models help clinicians identify patients who are more likely to benefit from specific treatment strategies while providing realistic expectations regarding treatment success (Tran et al., 2019; VerMilyea et al., 2020). Nevertheless, AI predictions should be interpreted alongside clinical judgment, because unexpected patient-specific factors may influence outcomes.
Embryo selection remains one of the most important determinants of IVF success. Conventional embryo grading depends largely on the experience of embryologists and may vary among fertility clinics or even among professionals within the same laboratory.
AI reduces this variability by applying standardized algorithms to evaluate embryo images. Unlike manual assessment, AI systems consistently apply the same criteria to every embryo, thereby improving objectivity and reducing observer bias. Time-lapse imaging combined with deep learning has further enhanced embryo evaluation by allowing continuous monitoring of embryo development without disrupting incubation conditions (Curchoe & Bormann, 2019; Khosravi et al., 2019).
Although many studies report improved prediction of embryo viability, evidence that AI consistently increases live birth rates across diverse clinical settings remains limited. Large multicenter clinical trials are still needed to confirm these findings.
AI contributes to the growing movement toward personalized medicine. Instead of applying the same treatment protocol to all patients, AI analyzes individual clinical characteristics to recommend tailored treatment plans.
Personalized ovarian stimulation protocols may improve the number of mature eggs retrieved while reducing complications such as ovarian hyperstimulation syndrome. AI may also identify patients with poor ovarian response or repeated implantation failure, allowing clinicians to modify treatment strategies earlier in the IVF process (ESHRE Guideline Group on Ovarian Stimulation, 2026).
This individualized approach has the potential to improve both clinical outcomes and patient satisfaction by reducing unnecessary treatment cycles.
Modern IVF laboratories generate large volumes of data through embryo imaging, semen analysis, laboratory monitoring, and electronic medical records. AI automates many routine tasks, reducing the workload of embryologists and laboratory personnel.
Automation improves workflow efficiency by reducing manual data entry, detecting equipment abnormalities early, standardizing laboratory procedures, supporting quality assurance and quality control, and minimizing human error. Improved laboratory efficiency may in turn allow fertility specialists to devote more time to patient counseling and complex clinical decision-making rather than repetitive administrative tasks.
Patients undergoing fertility treatment often experience considerable emotional and financial stress. AI-generated prediction models can provide individualized estimates of treatment success based on patient characteristics and previous clinical outcomes.
These estimates support informed discussions between clinicians and patients regarding expected outcomes, treatment options, and potential risks. More accurate counseling may improve patient understanding and help couples make informed reproductive decisions. Clinicians should, however, explain that AI predictions represent probabilities rather than guarantees of pregnancy.
AI has accelerated research in reproductive medicine by enabling the analysis of extensive clinical datasets collected from multiple fertility centers. Researchers can identify trends, evaluate treatment effectiveness, and develop improved predictive models more efficiently than with traditional statistical methods.
The increasing availability of international reproductive databases also supports collaborative research, facilitating the continuous improvement of AI algorithms through larger and more diverse patient populations.
Despite encouraging progress, the current evidence supporting AI in ART remains mixed. Many published studies report improved prediction accuracy, embryo ranking, and laboratory efficiency. However, relatively few large randomized controlled trials have demonstrated that AI consistently improves the most important clinical outcomes, including live birth rates.
Several limitations should be considered. First, many AI models are developed using retrospective data from individual fertility centers, which may limit their generalizability to different healthcare settings. Second, variations in patient demographics, laboratory protocols, imaging systems, and clinical practices may affect algorithm performance. Third, many commercial AI systems operate as “black boxes,” making it difficult for clinicians to understand how decisions are generated. This lack of transparency may reduce trust among healthcare professionals and patients.
Furthermore, algorithm performance depends heavily on data quality. Incomplete, inaccurate, or biased datasets can lead to unreliable predictions and may unintentionally disadvantage certain patient groups. Consequently, experts recommend the continuous validation of AI systems using diverse populations before routine implementation in clinical practice (ASRM, 2026).
In sum, the available evidence suggests that AI can improve efficiency, consistency, and decision support within ART. However, AI should currently be viewed as an adjunct to clinical expertise rather than an independent decision-maker. Continued clinical validation, transparent algorithm development, and rigorous regulatory oversight will be essential for ensuring its safe and effective integration into routine fertility care.
The growing use of artificial intelligence (AI) in assisted reproductive technology (ART) has created new opportunities for improving fertility care. It has also introduced complex ethical issues that extend beyond clinical effectiveness. While AI can facilitate decision-making and treatment efficiency, concerns remain regarding transparency, fairness, privacy, accountability, and equitable access. Addressing these ethical challenges is therefore essential to ensure that AI is implemented responsibly and in a manner that protects patients’ rights and promotes public trust.
Informed consent is a fundamental principle of medical ethics. Patients have the right to understand the benefits, risks, and limitations of any medical intervention before agreeing to treatment. The use of AI introduces additional complexity, because many patients may not fully understand how AI systems analyze data or generate recommendations.
Healthcare providers should therefore clearly explain the role of AI in fertility treatment, including how patient data will be used, the extent to which AI influences clinical decisions, and the limitations of algorithm-based predictions. Patients should also be informed that AI supports, rather than replaces, the professional judgment of fertility specialists. Providing clear and understandable information strengthens patient autonomy and promotes informed decision-making (Afnan et al., 2021).
AI systems learn from existing datasets. If these datasets do not adequately represent diverse populations, the resulting algorithms may produce biased results. For example, an AI model developed using data primarily from one ethnic group or geographical region may not perform equally well for patients from different backgrounds.
Algorithmic bias may result in unequal predictions of treatment success or inappropriate clinical recommendations for certain patient groups. Such disparities could widen existing inequalities in reproductive healthcare, especially among underserved populations. Developers should therefore train AI systems using large, diverse, and representative datasets while regularly evaluating algorithm performance across different demographic groups (WHO, 2024).
AI relies on large amounts of patient data, including medical histories, genetic information, laboratory results, and reproductive records. These data are highly sensitive and require strong protection against unauthorized access or misuse.
Maintaining confidentiality is essential, because reproductive information involves deeply personal aspects of an individual’s life. Fertility clinics should implement robust cybersecurity measures, encryption technologies, secure data storage, and strict access controls. Compliance with national and international data protection laws, such as the European Union’s General Data Protection Regulation (GDPR), is also necessary to safeguard patient privacy (European Union, 2016).
Patients should accordingly be informed about how their data will be collected, stored, shared, and used for both clinical care and future AI development.
Many AI systems operate as “black-box” models, meaning that clinicians and patients may not fully understand how the algorithm reaches its conclusions. This lack of transparency creates ethical concerns, because important reproductive decisions may be influenced by recommendations that cannot easily be explained.
Explainable AI has emerged as an important area of research aimed at improving transparency and clinician confidence. AI systems should provide understandable reasons for their recommendations whenever possible, allowing healthcare professionals to evaluate whether the outputs are clinically appropriate. Greater transparency also helps patients develop trust in AI-assisted healthcare while supporting accountability in clinical practice (Afnan et al., 2021; AI HLEG, 2019).
Determining responsibility for AI-assisted clinical decisions remains an important ethical challenge. If an AI system produces an inaccurate recommendation that contributes to treatment failure or patient harm, questions arise regarding who should be held accountable, whether the software developer, the healthcare institution, the fertility specialist, or the regulatory authority.
Current ethical guidance emphasizes that AI should function as a clinical decision-support tool rather than an independent decision-maker. Fertility specialists remain responsible for evaluating AI recommendations, discussing treatment options with patients, and making final clinical decisions. Human oversight is therefore essential to maintain patient safety and professional accountability (WHO, 2024).
Although AI has the potential to improve fertility care, access to advanced AI-assisted technologies remains uneven across countries and healthcare systems. High implementation costs, specialized infrastructure, and technical expertise may limit availability in low- and middle-income countries.
Unequal access raises concerns about justice and fairness. Patients in resource-limited settings may be unable to benefit from technological advances that improve treatment outcomes elsewhere. Policymakers and healthcare organizations should therefore develop strategies that promote equitable access to safe and effective AI technologies while avoiding further disparities in reproductive healthcare.
Reproductive medicine involves emotional, psychological, and social factors that cannot be fully addressed by AI. Couples seeking fertility treatment often require empathy, reassurance, and individualized counseling throughout their treatment journey.
While AI can provide valuable clinical insights, it cannot replace compassionate communication or the ethical judgment of experienced healthcare professionals. Fertility specialists must continue to consider patients’ personal values, cultural beliefs, financial circumstances, and reproductive goals when making treatment decisions. Maintaining a patient-centered approach ensures that technological innovation enhances rather than diminishes the quality of care.
In short, artificial intelligence offers significant opportunities to improve assisted reproductive technology, but its successful integration depends on careful attention to ethical principles. Issues relating to informed consent, algorithmic bias, privacy, transparency, accountability, and equitable access require ongoing collaboration among clinicians, researchers, policymakers, technology developers, and regulatory agencies. Establishing clear ethical guidelines and maintaining meaningful human oversight will help ensure that AI strengthens reproductive healthcare while protecting patient rights, promoting fairness, and sustaining public confidence in fertility medicine.
The rapid integration of artificial intelligence (AI) into assisted reproductive technology (ART) has challenged existing legal and regulatory systems. Most countries regulate fertility treatment and medical devices separately, while specific laws governing AI in reproductive medicine remain limited. As a result, many fertility clinics rely on general healthcare, data protection, and medical device regulations when implementing AI-based technologies. Developing clear and consistent regulatory frameworks is therefore essential to ensure patient safety, promote innovation, and maintain public confidence.
In the United States, ART is regulated through a combination of federal and state laws. The Food and Drug Administration (FDA) oversees certain AI-based medical devices under its medical device regulatory framework, while fertility practice guidelines are developed by professional organizations such as the American Society for Reproductive Medicine (ASRM) and the Society for Assisted Reproductive Technology (SART) (ASRM, 2026; U.S. Food and Drug Administration [FDA], n.d.).
AI systems intended for clinical decision support may require FDA review depending on their level of risk and intended use. However, many AI applications used in fertility clinics remain outside direct regulatory oversight because they function primarily as decision-support tools rather than autonomous medical devices. Consequently, professional standards and institutional policies continue to play an important role in ensuring safe implementation.
Canada regulates reproductive medicine through the Assisted Human Reproduction Act (S.C. 2004, c. 2) and related provincial healthcare laws. While the Act provides a legal framework for assisted reproduction, it was enacted before the widespread use of AI and therefore contains no provisions addressing AI-assisted clinical decision-making.
Health Canada regulates medical devices, including software-based technologies that meet the definition of a medical device. AI systems used for diagnosis or treatment recommendations may therefore require regulatory approval before clinical use. In addition, Canadian privacy legislation requires healthcare organizations to protect patients’ personal and health information throughout data collection, storage, and analysis (Health Canada, 2026).
As AI adoption increases, Canadian regulators are considering broader governance strategies that emphasize transparency, safety, and responsible innovation.
The United Kingdom has one of the world’s most established regulatory systems for assisted reproduction. The Human Fertilisation and Embryology Authority (HFEA) oversees fertility clinics, embryo research, and ART procedures under the Human Fertilisation and Embryology Act 1990, as amended in 2008.
Although the legislation does not specifically regulate AI, the HFEA has recognized the growing importance of AI in embryo assessment and fertility treatment. The Authority supports innovation while emphasizing that AI technologies should undergo rigorous scientific validation before routine clinical use. AI-assisted decisions must remain subject to professional oversight, and clinics are expected to demonstrate that new technologies improve patient outcomes without compromising safety or ethical standards (HFEA, 2026).
The European Union has taken a comprehensive approach to AI regulation through the Artificial Intelligence Act, which establishes risk-based rules for AI systems used across different sectors, including healthcare. Under the Regulation, AI systems that are medical devices, or safety components of medical devices, subject to third-party conformity assessment are classified as high-risk and must comply with strict requirements relating to safety, transparency, human oversight, risk management, and data quality (European Union, 2024).
The General Data Protection Regulation (GDPR) also provides strong protection for personal health information by requiring lawful data processing, informed consent where appropriate, and safeguards against the unauthorized use of sensitive personal data (European Union, 2016). Together, these regulations provide one of the most comprehensive legal frameworks for AI in healthcare.
Australia regulates assisted reproduction primarily through state and territory legislation, supported by national ethical guidelines developed by the National Health and Medical Research Council (NHMRC) (NHMRC, 2023). AI technologies that function as medical devices are regulated by the Therapeutic Goods Administration (TGA).
Australian regulators encourage innovation while emphasizing clinical validation, patient safety, transparency, and responsible data management. Fertility clinics introducing AI-based systems are expected to maintain appropriate quality assurance procedures and ensure that clinical decisions remain under the supervision of qualified healthcare professionals.
Nigeria’s use of ART has expanded considerably during the past decade, particularly through private fertility clinics. However, the country does not yet have a comprehensive legal framework specifically regulating AI in reproductive medicine.
Healthcare providers currently rely on general medical ethics, professional practice standards, and data protection legislation, including the Nigeria Data Protection Act 2023, to guide the handling of patient information. While these laws provide important safeguards for privacy and confidentiality, they do not specifically address issues such as algorithm validation, AI accountability, or automated clinical decision-making.
As AI adoption continues to increase, Nigeria has an opportunity to develop national guidelines that address the safe use of AI in fertility care. Such regulations could establish standards for clinical validation, patient consent, cybersecurity, algorithm transparency, and professional responsibility while encouraging responsible technological innovation.
Despite differences among national legal systems, several common regulatory principles are emerging. These include protecting patient safety, ensuring transparency, maintaining human oversight, safeguarding personal data, validating AI systems through clinical research, and establishing clear accountability for AI-assisted decisions.
Greater international collaboration among governments, professional societies, regulatory agencies, and technology developers would help promote consistent standards for AI in reproductive medicine. Harmonized regulations could facilitate responsible innovation while ensuring that patients receive safe, effective, and ethically sound fertility care regardless of where treatment is provided.
In summary, the legal regulation of AI in assisted reproductive technology is evolving but remains fragmented across jurisdictions. Countries such as the United Kingdom and the member states of the European Union have developed relatively advanced governance frameworks, whereas other jurisdictions, including Nigeria, are still in the early stages of regulatory development. As AI becomes increasingly integrated into fertility practice, governments and professional organizations will need to update existing laws to address emerging issues relating to patient safety, transparency, privacy, accountability, and equitable access. A balanced regulatory approach that encourages innovation while protecting patients will be essential for the responsible future of AI-assisted reproductive medicine.
Based on the findings of this review, the following recommendations are proposed to support the responsible integration of artificial intelligence (AI) into assisted reproductive technology (ART):
1. Strengthen clinical validation. AI systems should undergo rigorous multicenter clinical trials involving diverse patient populations before they are widely adopted in fertility practice. Continuous post-implementation monitoring should also be conducted to evaluate their long-term effectiveness and safety.
2. Maintain human oversight. AI should serve as a clinical decision-support tool rather than replace fertility specialists. Final treatment decisions should always remain the responsibility of qualified healthcare professionals who can consider each patient’s unique medical and personal circumstances.
3. Develop comprehensive regulatory frameworks. Governments and regulatory agencies should establish clear legal guidelines for the approval, monitoring, and use of AI in reproductive medicine. These regulations should address algorithm validation, accountability, transparency, patient safety, and ethical standards.
4. Improve data privacy and cybersecurity. Fertility clinics should implement strong data protection measures, including secure data storage, encryption, restricted access, and compliance with applicable privacy laws. Patients should be fully informed about how their personal and reproductive health data will be collected, stored, and used.
5. Promote algorithm transparency. Developers should prioritize explainable AI systems that enable clinicians to understand how recommendations are generated. Transparent algorithms will improve clinician confidence, facilitate patient communication, and enhance accountability.
6. Reduce algorithmic bias. AI models should be trained using diverse and representative datasets to minimize bias and improve accuracy across different populations. Regular auditing should be conducted to identify and correct potential disparities in algorithm performance.
7. Expand professional training. Fertility specialists, embryologists, and other healthcare professionals should receive regular training on AI technologies, their capabilities, limitations, and ethical implications to ensure their responsible use in clinical practice.
8. Promote equitable access. Policymakers should develop strategies that improve access to AI-assisted fertility care in low- and middle-income countries through investment in healthcare infrastructure, international collaboration, and affordable technology solutions.
9. Encourage interdisciplinary collaboration. The future development of AI in ART should involve collaboration among clinicians, computer scientists, ethicists, legal experts, policymakers, and patient representatives to ensure that technological innovation aligns with clinical needs and ethical principles.
10. Support future research. Further studies should evaluate the long-term impact of AI on live birth rates, patient satisfaction, cost-effectiveness, and healthcare equity. Comparative studies across different healthcare systems will also help identify best practices for AI implementation.
Artificial intelligence is reshaping assisted reproductive technology by improving clinical decision-making, laboratory efficiency, and personalized fertility care. Current applications, including embryo selection, IVF outcome prediction, sperm analysis, ovarian stimulation planning, and laboratory automation, demonstrate the potential of AI to enhance the accuracy, consistency, and efficiency of fertility treatment. These developments may contribute to improved pregnancy outcomes while supporting clinicians in managing increasingly complex reproductive data.
Despite these promising advances, AI is not without challenges. Ethical concerns relating to informed consent, algorithmic bias, transparency, data privacy, accountability, and equitable access remain significant barriers to its widespread adoption. Similarly, existing legal and regulatory frameworks vary considerably across countries, creating uncertainty regarding the governance and safe implementation of AI-assisted reproductive technologies. Addressing these challenges will require continuous collaboration among healthcare professionals, technology developers, researchers, regulators, and policymakers.
Current evidence indicates that AI should complement rather than replace the expertise and clinical judgment of fertility specialists. Human oversight remains essential to ensure that treatment decisions consider not only algorithmic predictions but also patients’ medical histories, personal values, emotional well-being, and individual reproductive goals.
Looking ahead, advances in machine learning, deep learning, genomic medicine, and digital health are expected to expand the role of AI in reproductive medicine further. The successful integration of these technologies will, however, depend on rigorous clinical validation, transparent algorithm development, robust regulatory oversight, effective data governance, and adherence to ethical principles that protect patient rights and promote public trust.
In conclusion, artificial intelligence represents one of the most significant technological innovations in modern reproductive medicine. When implemented responsibly within an appropriate ethical and regulatory framework, AI has the potential to improve the quality, safety, accessibility, and personalization of assisted reproductive technology while contributing to better reproductive health outcomes worldwide.
*****