| Name |
Risk of AI Development Impacting Decision-making Accuracy |
Risk of Disinformation Spread Undermining Public trust and Communications Trust |
| Risk Category |
Technological |
Technological |
| Description |
- With the rapid evolution of generative AI, large language models, and automated decision-making technologies, the complexity, black-box nature, and reliance on external data sources of AI models continue to increase, potentially leading to reduced decision explainability, model bias, unstable outputs, and risks concerning data licensing, privacy, and security.
- As AI regulatory frameworks and industry governance standards are still developing, external regulatory requirements, technical standards, and novel attack vectors are changing rapidly, potentially exposing enterprises to higher governance uncertainties when deploying AI applications.
- Involving technology, cybersecurity, compliance, ethics, and operational management, the risk constitutes an emerging risk of technology with long-term impacts.
|
- Misinformation and disinformation, amplified by social media algorithms, are prone to creating widespread public opinion turbulence. Such content often blends partial truths with AI-generated imageries, fabricated sources, or decontextualized information, making it difficult to identify and clarify, potentially undermining user trust in Chunghwa Telecom's ICT services.
- As a provider of critical communication services and digital interfaces, if Chunghwa Telecom 's brand, services, or infrastructure are falsely associated with misinformation, impersonated, or maliciously manipulated, it may raise user concerns regarding its service stability, cybersecurity, and corporate credibility, resulting in burdens on customer service, clarification, and crisis response. Driven by rapid changes in the external information environment and technological tools, the risk presents long-term, cross-domain, and highly uncertain impacts.
|
| Impact |
- Reallocating governance budgets and tightening reviews resulting from deployment delay due to risk control, AI governance, and regulatory demands for operational AI integration and universal employee adoption may impact market expansion, efficiency, and profitability.
- Failing to systematically detect and correct AI model bias may result in, process deviations, infringement issues, and compliance disputes that damage brand trust, requiring costly remediation and increasing operational pressure.
- Rising societal concerns over AI ethical risks will drive stricter regulations, escalating compliance costs, internal control demands, and technical governance burdens, thus limiting AI innovation and commercialization.
- Increasing AI system complexity elevates demands for model maintenance, risk monitoring, AI governance, and tech updates, resulting in a surge in operational expenditures that may impact free cash flow, investment flexibility, and long-term enterprise value.
|
- The rapid spread of disinformation and deepfakes on social media may cause stakeholders to misjudge services, cybersecurity incidents, or financials, leading to investment errors, complaint surges, abnormal user behavior, and unfair competition with impacts on operational efficiency, service trust, brand reputation, and market expansion.
- The Company must increase resources for public opinion monitoring, fact-check, crisis response, clarification, customer service, and legal consultation, driving operational costs and affecting ESG ratings.
- Failure to promptly identify, clarify, and respond lowers trust from governments, corporate clients, and partners, hindering service innovation. Incidents involving fraud, fictitious financials, or false investment information could trigger stock volatility, administrative fines, legal disputes, and liabilities, adversely affecting investor confidence, customer retention, and long-term enterprise value and elevating overall business risks.
|
| Mitigating actions |
- Chunghwa Telecom has established its AI policy, organization, and systems in accordance with ISO 42001 and the Artificial Intelligence Basic Act referencing competent authority AI guidelines.
- For LLM applications, Chunghwa Telecom has deployed risk management processes, integrating AI security assessments, output validation, model monitoring, and data protection to enhance the explainability, accuracy, and compliance.
- Partnering with industry, government, and academia, Chunghwa Telecom promotes AI education to bolster public trust, conducts tiered training on AI risks and cybersecurity, and tracks completion and pass rates, continuously strengthening its risk response capabilities.
- Integrate AI technology applications with international AI standards to build a "technical protection-architectural resilience-resource allocation" 3-tiered management framework, enhancing risk protection, architectural flexibility, and resource control governance.
|
- Implement AI technology to build an automated monitoring framework to identify potential controversial issues and unusual information dissemination in real time.
- Integrate multi-source data platforms for real-time social and media opinion capture and semantic analysis to grasp issue trends.
- Utilize machine learning to optimize semantic classification models, improving the labeling and grading.
- Integrate NLP to enhance the ability to analyze potential controversy.
- Establish a model refinement mechanism with regular updates of training datasets and algorithm in response to the advancement of misinformation.
- Internal Crisis Response Planning:
(1). Strengthen the coordinated handling capability among strategy, customer service, public relations, and marketing departments.
(2). Monitor high-risk overseas social media platforms to enable early warning and risk prevention.
|
