Table 3: Challenges, risks, and mitigation strategies for AI and digital health in medical education.

Challenge Domain

Description

Proposed Mitigation Strategies

Urgency

Key References

Content accuracy and hallucination

LLMs may generate plausible but factually incorrect content, risking propagation of misinformation

Mandatory expert review; retrieval-augmented generation; iterative validation; transparent disclosure of limitations

High

[12,22]

Algorithmic bias

Models may reproduce racial, gender, or socioeconomic biases present in training data, affecting vignettes and recommendations

Diverse training datasets; regular bias audits; inclusive design teams; equitable representation

High

[13,20,59]

Data privacy and security

Student performance data, biometric data from wearables, and simulated patient data raise privacy concerns

GDPR/HIPAA compliance; anonymization; institutional review; transparent data governance

High

[16,17]

Digital divide

Unequal access to technology, connectivity, and infrastructure limits participation of learners in LMICs

Low-bandwidth solutions; mobile-first design; institutional investment; international collaboration and funding

High

[9,23]

Faculty readiness

Many educators lack training in AI and digital health, hindering effective curricular integration

Faculty development programs; communities of practice; mentorship; institutional incentives

Medium

[1,24]

Academic integrity

LLMs enable dishonesty in assignments, assessments, and scholarship, challenging traditional evaluation

Detection tools; redesigned assessments emphasizing reasoning; proctoring; honor codes

Medium

[2,19,60]

Over-reliance on technology

Excessive dependence on AI may diminish independent reasoning, empathy, and humanistic skills

Balanced curriculum design; emphasis on clinical exposure; reflective practice; human–AI collaboration

Medium

[5,21]

Regulatory and governance gaps

Lack of standardized guidelines for AI use; variable institutional policies across regions

National/international frameworks; accreditation standards; ethical AI governance committees

Medium

[13,25]

Note: Urgency levels were categorized as High (requiring immediate attention), Medium (important but less time-sensitive), or Low (long-term consideration). GDPR = General Data Protection Regulation; HIPAA = Health Insurance Portability and Accountability Act; LMICs = Low- and Middle-Income Countries.