To ensure transparency and trustworthiness throughout the rollout of medical AI—and thereby boost end‑user confidence in clinical settings—China Medical University Hospital (CMUH) established the Smart Healthcare Committee under the guidance and funding of the Ministry of Health and Welfare’s 2024 (Year 113) Next‑Generation Digital Healthcare Platform: Three AI Centers Grant Program. After the grant period, this committee will operate permanently as CMUH’s Responsible AI Implementation Center.

AI Development Declaration & Action Guidelines

On 2024‑07‑15, the National Science and Technology Council released a draft Fundamental Act of Artificial Intelligence, defining seven AI values: Safety, Accountability, Fairness, Transparency, Privacy, Sustainability, Autonomy.

1. Sustainability & Well‑being: Promote social equity and environmental sustainability. Provide proper education and training to narrow the digital divide and help citizens adapt to AI‑driven change.
2. Human Autonomy: Support human autonomy and dignity, respect fundamental rights and cultural values, and allow human oversight—upholding a people‑centric, rule‑of‑law, democratic approach.
3. Privacy Protection & Data Governance: Safeguard personal‑data privacy, mitigate data‑leak risks, apply data‑minimization, and encourage openness and reuse of non‑sensitive data.
4. Cybersecurity & Safety: Throughout AI R&D and deployment, establish security measures to defend against threats and attacks, ensuring system robustness and safety.
5. Transparency & Explainability: Disclose or label AI outputs appropriately so stakeholders can assess potential risks and understand impacts—thereby strengthening trust.
6. Fairness & Non‑discrimination: Minimize algorithmic bias and discriminatory risks; AI must not produce discriminatory outcomes for specific groups.
7. Accountability: Assume both internal governance responsibility and external social responsibility.

In this spirit, CMUH issues the following AI Development Declaration:

Sustainability & Well‑being

We will deepen AI education and training for staff, expanding from medical AI to administrative AI.

Human Autonomy

We will develop patient‑centric medical AI that benefits humanity and actively pursue FDA and TFDA approvals.

To implement these goals—and leveraging our existing HIMSS and ISO certifications—CMUH sets the following AI Action Guidelines:

Privacy Protection & Data Governance

Through HIMSS AMAM and ISO 27701/29191 certification, we will strengthen privacy protection and data governance in AI.

Cybersecurity & Safety

Through HIMSS INFRAM and ISO 27001 certification, we will reinforce cybersecurity defenses for AI and guard against threats and attacks.

Transparency & Explainability

Following the nine Transparency Disclosure mechanisms, we will ensure medical AI is transparent and explainable, publishing relevant information on the CMUH website.

Fairness & Non‑discrimination

We will apply the nine Transparency Disclosure mechanisms to guarantee fairness and prevent algorithmic bias.

Accountability

In line with the Medical‑AI Implementation Regulation, we will responsibly manage trial/production applications and life‑cycle monitoring of medical AI.

Three Core Strategies for Responsible‑AI Implementation