Text‐to‐Video Generative Artificial Intelligence in Pediatrics: Potential Applications, Ethical Considerations, and Future Directions

ABSTRACT

Background

Text‐to‐video (T2V) generative artificial intelligence (AI) models can produce short videos from natural‐language prompts, creating new opportunities for medical visualization and communication. In pediatrics, visual explanations and preparatory media are already used to support learning and to reduce distress during procedures; however, current T2V systems (including proprietary models such as OpenAI's Sora) have not been clinically validated for routine pediatric care.

Objective

To provide a forward‐looking perspective on how T2V generative AI could be used in pediatric education, patient and caregiver communication, and simulation, while clearly distinguishing current evidence from hypothetical applications.

Approach

We performed a narrative (nonsystematic) review of the literature on (i) generative AI and T2V models in healthcare and (ii) pediatric video‐based education, procedural preparation, and digital distraction (PubMed and Google Scholar; search through June 2025; representative keywords included “pediatric” AND “video‐based education,” “procedural preparation,” “anxiety,” “pain,” and “simulation”).

Key Points

Potential applications include scenario‐based training videos for clinicians, tailored educational animations for families, and child‐friendly preparatory videos to reduce anxiety and pain. Key limitations and risks include inaccurate or oversimplified content, bias, privacy and data‐governance concerns, unclear accountability, commercialization pressures, and inequitable access. Pediatric‐specific safeguards (transparent disclosure, clinician review, age‐appropriate design, consent/assent practices, and equity planning) are essential.

Conclusion

T2V generative AI should be framed as an experimental adjunct rather than a near‐term clinical solution. Rigorous validation, governance, and equity‐focused implementation strategies are needed before pediatric deployment.

1. Introduction

Text‐to‐video (T2V) generative AI refers to models that generate short videos from natural‐language prompts. OpenAI's Sora is one prominent proprietary example in this rapidly evolving class of systems. In healthcare, the clinical value of any T2V model depends on validated accuracy, safety, and governance – requirements that are not yet met for routine pediatric use.

Technically, T2V systems often combine large language models (LLMs) for prompt understanding with diffusion‐based generation to synthesize temporally coherent video content. This capability could be useful for medical visualization and communication, but current outputs may contain hallucinated details, bias, or clinically misleading simplifications.

Literature approach (brief): This article is a narrative perspective. We identified representative publications on (a) generative AI/T2V models in healthcare and (b) pediatric video‐based education and procedural preparation. We prioritized peer‐reviewed studies and systematic reviews, and we use Sora only as an illustrative case rather than as an endorsed product.

2. Pediatric Education

The use of LLMs also raises ethical concerns. Potential risks include biases in generated content, plagiarism, and a possible shift away from face‐to‐face education. To mitigate these risks, it is crucial to implement regulatory mechanisms that minimize misinformation and bias in LLM‐generated outputs. Additionally, educational modules should be developed to help students understand both the advantages and limitations of this technology. When applied cautiously and systematically, the benefits of LLMs in pediatric education can outweigh their risks [1].

Video‐based learning approaches have been notably effective in pediatric education. Studies indicate that participation in video case analysis programs enhances medical students' pediatric patient assessment skills [2]. These programs also improve clinicians' confidence in pediatric interventions and promote adherence to evidence‐based practices in their clinical routines [3]. By integrating with such technologies, Sora could contribute to setting new standards in pediatric education. For instance, a simulation illustrating a child experiencing respiratory distress can help students learn how to manage such a situation. Interactive and realistic simulations like these not only boost student confidence but also refine their clinical application skills.

Importantly, no peer‐reviewed pediatric studies have validated T2V‐generated clinical content; therefore, any educational use should be considered exploratory and subject to expert review.

3. Visualization and Patient Education

The application of generative AI in pediatrics holds significant importance due to its impact on the way physicians educate families about childhood illnesses and its influence on physician–patient and physician–caregiver relationships [4]. Research suggests that generative AI is effective in supporting therapy processes, facilitating emotional health assessments, and providing personalized solutions for children with special needs [5]. Moreover, generative AI's T2V models have transformative potential in patient education, medical training, and telemedicine.

Emerging T2V platforms can generate high‐quality video content, offering innovative solutions for healthcare services. For example, Sora's animation capabilities can be utilized to comprehensively explain vaccination procedures or treatment protocols to alleviate parental concerns. By visually simplifying medical procedures, such videos can enhance parental awareness and support informed decision‐making based on accurate and reliable information. Furthermore, these platforms can provide interactive videos aimed at reducing anxiety in children undergoing medical treatments. For instance, Sora could generate videos simulating a blood draw procedure in an engaging and child‐friendly manner, helping children adapt more comfortably to the process.

However, challenges such as misinformation, privacy violations, and ethical dilemmas persist. Thus, the development of detection mechanisms and ethical frameworks is necessary for the responsible and effective implementation of these technologies [6].

Existing pediatric evidence supports the general concept that child‐friendly videos can improve understanding and reduce anxiety in diagnostic or procedural contexts (e.g., video preparation before MRI and short video interventions before procedures), and that digital distraction can reduce procedural pain and anxiety. These data provide indirect support for exploring customized visual media, but they do not validate AI‐generated video content specifically [7, 8, 9, 10, 11]. Potential pediatric applications, expected benefits, key risks, and current evidence level are summarized in Table 1.

Table 1.

Potential pediatric applications of text‐to‐video generative AI: benefits, risks, and current evidence level.

Application domain	Example use‐case	Potential benefits	Key risks	Current evidence level
Clinician education and simulation	Scenario videos for rare pediatric emergencies from standardized scripts	Scalable visualization; standardization; debriefing support	Incorrect steps; bias; false confidence; IP issues	Conceptual; extrapolated
Patient/caregiver education	Age‐tailored animations explaining conditions/procedures	Improved comprehension; engagement	Oversimplification; misinformation; opaque provenance; liability	Indirect support (pediatric video education)
Anxiety/pain reduction	Preparatory or distraction videos before venipuncture, imaging, endoscopy	Reduced distress; potentially less restraint/sedation	Manipulative content; inequitable access; triggering imagery	Indirect support (RCTs/meta‐analyses)
Neurodevelopmental communication	Social stories/step‐by‐step sequences for ASD/ADHD	Predictability; reduced sensory distress	Inappropriate framing; privacy risks; overreliance	Limited; not specific to T2V AI
Public health messaging	Short explainer videos for vaccination/prevention	Consistent messaging; wide reach	Misinformation amplification; trust erosion	Evidence for video messaging; not for T2V AI

4. Future Perspectives and Areas of Innovation

Advanced generative AI models like Sora hold groundbreaking potential in critical areas of pediatrics, including patient education, psychological support, and visualization of treatment processes. Furthermore, iterations of Sora and other T2V generative AI applications have been suggested to offer significant benefits to the neurosurgical community [12]. Sora's ability to create realistic and creative scenes from textual commands makes it a valuable tool for patient education, clinical simulations, and personalized treatment processes [13].

For example, detailed animations explaining vaccination procedures or surgical steps could help reduce parental anxiety regarding medical processes. Additionally, interactive and engaging content could facilitate pediatric patients' adaptation to treatments by mitigating anxiety. Sora can explain stressful procedures, such as blood draws, in a visually appealing and child‐friendly manner, enhancing pediatric patients' compliance with medical interventions.

Despite the promising potential of AI‐generated videos in enhancing patient education, facilitating remote consultations, and simulating surgical procedures, these technologies also present technical, legal, and ethical challenges [14].

To ensure the effective integration of these technologies into healthcare services, it is essential to develop regulatory mechanisms, establish ethical guidelines, and implement fair policies to address accessibility disparities. These measures will help maximize the positive impact of Sora on pediatric healthcare and expand its benefits to a broader population.

5. Ethical, Regulatory, and Equity Considerations

Ethical analysis in pediatrics must extend beyond general AI concerns. Children are a vulnerable population, and visual content can be especially persuasive. When AI‐generated videos are used for education or preparation, clinicians should disclose that the content is AI‐generated, explain its limitations, and ensure that information aligns with current clinical standards.

5.1. Consent and Assent

Caregivers should provide informed consent for the use of AI‐generated educational materials, and children should receive age‐appropriate explanations. Content should avoid fear‐inducing or emotionally manipulative imagery and should respect developmental stage and cultural context.

5.2. Accountability and Safety

A clear human‐in‐the‐loop review process is required, with responsibility assigned for clinical accuracy, updates, and incident response (e.g., correction of detected misinformation). Privacy and data governance should prohibit sharing identifiable pediatric data in prompts, and organizations should evaluate vendor terms, data retention, and security controls.

5.3. Equity

Reliance on high‐bandwidth digital tools may widen disparities for families with limited access to devices, connectivity, or digital literacy. Implementation plans should include low‐tech alternatives, accessibility features, multilingual options, and evaluation across diverse populations.

5.4. Commercial Neutrality

Sora is used only as a case example. The discussion focuses on the broader class of T2V models, and any deployment should be guided by transparent procurement, conflict‐of‐interest management, and independent evaluation.

6. Conclusion

In conclusion, the integration of generative AI technologies like Sora into pediatric healthcare may have the potential to support patient education and treatment processes, making healthcare services more efficient and innovative. However, to fully realize this potential, a careful and ethical approach must be adopted. Generative AI should be regarded as a complementary tool to human expertise in healthcare, and robust ethical frameworks must be established to minimize risks such as misinformation and privacy breaches. Additionally, fostering interdisciplinary collaboration among stakeholders, healthcare professionals, and technology developers is crucial to ensuring the fair and accessible implementation of this technology. Further research, development, and policy‐focused efforts are necessary to enhance Sora's contributions to pediatric healthcare.

Author Contributions

Emine Ozdemir Kacer: investigation, writing – original draft, conceptualization, methodology, visualization, writing – review and editing, supervision, formal analysis.

Funding

The author received no specific funding for this work.

Disclosure

The lead/corresponding author Emine Ozdemir Kacer affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

Ethics Statement

This study is based solely on a review of existing literature and theoretical analyses; therefore, ethical approval was not required.

Conflicts of Interest

The author declares no conflicts of interest.

Data Availability Statement

Data available within the article.