Responsible Artificial Intelligence
LITEON is committed to developing and applying Artificial Intelligence (AI) in a responsible, transparent, secure, and ethical manner. While leveraging AI to drive innovation, improve operational efficiency, and strengthen organizational capabilities, LITEON actively manages risks related to data privacy, cybersecurity, algorithmic bias, content accuracy, transparency, and accountability.
To ensure that AI applications align with corporate governance, regulatory compliance, and sustainability objectives, LITEON has established an Artificial Intelligence Governance Policy covering the use and development of AI tools, training data management, AI model lifecycle management, and AI-generated content management. Through governance mechanisms, access controls, employee training, AI adoption initiatives, and incident reporting processes, LITEON continuously advances its Responsible AI program.
Artificial Intelligence Governance Policy
Scope and Governance Responsibilities
The Artificial Intelligence Governance Policy applies to all business units, functional departments, employees, suppliers, and relevant stakeholders across LITEON. It covers the use and development of AI tools, AI training data management, AI model management, and AI-generated content management.
The Information Security & Information Technology Center (IS&IT Center), led by the Chief Information Security Officer (CISO), serves as the AI governance function responsible for maintaining AI governance requirements, implementing information security controls, and reviewing and approving AI tool usage requests. AI development teams are responsible for the design, development, training, and testing of AI systems, while users remain accountable for validating AI-generated outputs and making final decisions.
Board Oversight
AI governance is incorporated into LITEON’s corporate governance framework. The Artificial Intelligence Governance Policy was approved by the Chairman and reviewed by the Board-level Corporate Sustainability Development Committee before being reported to the Board of Directors.
The Board oversees AI-related risks and opportunities through existing governance mechanisms, including information security, data protection, regulatory compliance, ethical considerations, environmental impacts, and responsible AI deployment.
Data Privacy and Cybersecurity
Protecting confidential business information and personal data is a core principle governing AI usage at LITEON. All AI-related activities must comply with the Company’s information security and data protection requirements to ensure data integrity, accuracy, confidentiality, and security.
LITEON also implements safeguards against AI-specific cybersecurity risks, including prompt injection, model poisoning, model extraction, and adversarial attacks. Logging, traceability, and incident reporting mechanisms are managed in accordance with the Company’s Information Security Policy.
Fairness, Human Oversight and Accountability
LITEON recognizes that AI systems may generate inaccurate, biased, or hallucinatory outputs due to data quality, model design, or inappropriate usage. Users are required to validate AI-generated outputs using professional judgment and appropriate review procedures before use.
LITEON applies a Human-in-the-Loop approach to critical decision-making processes, ensuring that final decisions remain subject to human review, approval, modification, or rejection by authorized personnel. Accountability for final outcomes remains with the responsible decision-maker.
Transparency, Explainability and AI Content Labeling
AI-generated or AI-assisted text, images, and multimedia content used in public communications or external engagements are identified through appropriate disclosures, enabling stakeholders to recognize AI involvement in content creation.
AI Usage Boundaries and Prohibited Applications
All AI tools must undergo internal review, approval, and access control procedures before deployment. Non-approved third-party AI platforms may not be used to process confidential business information, personal data, or other restricted information.
LITEON does not permit the use of AI in ways that violate applicable laws, information security requirements, privacy rights, intellectual property rights, or fundamental human rights.
The Company prohibits the use or deployment of AI applications that are classified as prohibited under applicable regulations. LITEON continuously monitors developments in the EU AI Act and other relevant regulations and international standards and updates its AI governance requirements accordingly.
Environmental Impact Management
LITEON recognizes the environmental impacts associated with AI computing, AI models, and data center operations. When developing, adopting, or deploying AI systems and services, LITEON considers computational efficiency and energy consumption alongside functional, cybersecurity, data protection, and regulatory requirements.
LITEON supports the development of energy-efficient AI infrastructure through high-efficiency power solutions, energy management capabilities, and system integration technologies. The Company also continues to monitor emerging technologies, computing architectures, renewable energy applications, and lower-impact AI solutions as part of its broader sustainability commitments.
Responsible Artificial Intelligence Program
AI Governance, Access Control and Model Management
LITEON provides approved AI tools through controlled and secure environments to reduce cybersecurity and compliance risks associated with unauthorized platforms.
AI model management covers the full lifecycle from requirement definition and data collection through development, testing, deployment, monitoring, and retirement. Version control, logging, and traceability mechanisms support risk management, investigations, and continuous improvement.
Bias Review and Human Validation
Potential risks associated with AI-generated outputs, including inaccuracies, hallucinations, and bias, are managed through validation procedures, human review, model version management, and governance controls. AI outputs are used to support, rather than replace, professional judgment and decision-making.
AI Training and Capability Building
Employees using AI tools are required to complete relevant training and understand associated risks and governance requirements. LITEON provides training programs covering AI governance, foundational AI knowledge, approved AI tools, Microsoft Copilot, automation technologies, AI agents, and practical AI applications.
Through learning platforms, AI capability-building initiatives, AI champions, and cross-functional projects, LITEON continues to integrate AI competencies into research and development, manufacturing, and business operations.
Quantified Benefits and Operational Improvements
LITEON applies AI, Generative AI, and automation technologies across manufacturing, quality management, supply chain, engineering, and operational processes. Performance indicators are used to monitor improvements in process efficiency, operational effectiveness, data quality, and resource utilization.
Existing use cases have demonstrated productivity improvements exceeding 65%, with selected processes achieving efficiency gains of more than 90%, while reducing repetitive manual work and paper-based activities. LITEON continues to monitor measurable outcomes to support process optimization, operational excellence, and continuous improvement.
Supporting Energy-Efficient AI Infrastructure
As demand for AI, high-performance computing, and cloud services continues to grow, LITEON invests in high-efficiency server power supplies, data center power solutions, and advanced cooling technologies to improve power conversion efficiency, reduce energy losses, and optimize energy management for AI computing environments.
LITEON has showcased an 800 VDC Power Rack integrated with liquid-cooling technology to support emerging AI Factory requirements, including high-density computing, advanced cooling, and high-voltage DC power distribution. Through continuous innovation in power electronics, energy management technologies, and system integration capabilities, LITEON supports the development of more energy-efficient and lower-carbon AI infrastructure.
In addition, LITEON continues to promote renewable energy adoption and energy management initiatives, increasing the use of low-carbon energy sources while helping improve the energy efficiency of AI infrastructure and data center operations.
Incident Reporting, Review and Continuous Improvement
Logging, traceability, and incident reporting related to AI systems are managed in accordance with the Company’s Information Security Policy. Employees may report concerns involving data leakage, unauthorized access, inaccurate outputs, potential bias, or other AI-related issues through established reporting mechanisms.
Reported issues involving AI-generated outputs or AI-assisted decisions are subject to human review and appropriate corrective actions. Findings are incorporated into risk assessments, governance enhancements, and continuous improvement activities.
