content area
KBS AI Guidelines
Principles for Responsible Use of AI
Artificial intelligence (AI) technology serves as a valuable auxiliary tool, capable of significantly enhancing the quality of services and content that public service broadcaster KBS provides to the public, alongside boosting production efficiency. Having proactively adapted to changes in the broadcasting technology landscape, KBS is committed to further upholding its public service mission by actively harnessing AI technology. However, current AI technology also presents inherent risks, potentially giving rise to legal and ethical issues. Therefore, as a trusted leading public service broadcaster in Korea, KBS has established these guidelines to clearly define the principles and standards for the responsible use of AI technology.
These guidelines present fundamental principles and standards that KBS must adhere to for the responsible and ethical application of AI technology in the course of fulfilling its public service mission. The principles and standards outlined in these guidelines apply to all public services and content provided by KBS, and to all stakeholders involved in their production and operation.
Herein, 'Artificial Intelligence (AI)' refers to technology that electronically embodies human intellectual capabilities such as learning, reasoning, perception, judgment, and language comprehension, encompassing machine learning, natural language processing, and content generation tools.
These guidelines present fundamental principles and standards that KBS must adhere to for the responsible and ethical application of AI technology in the course of fulfilling its public service mission. The principles and standards outlined in these guidelines apply to all public services and content provided by KBS, and to all stakeholders involved in their production and operation.
Herein, 'Artificial Intelligence (AI)' refers to technology that electronically embodies human intellectual capabilities such as learning, reasoning, perception, judgment, and language comprehension, encompassing machine learning, natural language processing, and content generation tools.
- Article 1 (Human-Centered)
- The utilization of Artificial Intelligence (AI) must be conducted with human-centered values as the foremost priority.
- KBS's public services and content production using AI must contribute to enhancing human creativity and operational efficiency.
- AI must be utilized as a tool to support human autonomy and decision-making, and its application, from design, development, and operation to the generation of results, must undergo continuous, multi-layered human oversight and final approval throughout the entire process.
- The utilization of Artificial Intelligence (AI) must be conducted with human-centered values as the foremost priority.
- Article 2 (Public Service Mission)
- The utilization of AI must align with the values pursued by public service broadcaster KBS.
- Public services and content production utilizing AI must conform to KBS's internal standards, including the 'KBS Broadcast Production Guidelines'.
- Due diligence must be exercised regarding the inherent risk of bias in AI, and appropriate measures must be taken to prevent it.
- In the process of utilizing AI, diverse societal values and interests must be evenly reflected, political and religious impartiality maintained, and vulnerable groups protected.
- When utilizing AI in news and current affairs content, particular consideration must be given to fairness and accuracy.
- The utilization of AI must align with the values pursued by public service broadcaster KBS.
- Article 3 (Accountability)
- When utilizing AI, it is important to acknowledge that legal and ethical responsibilities are inherent, and careful attention must be paid to prevent issues.
- Relevant laws and regulations must be observed when utilizing AI.
- Public services and content produced with AI must undergo final verification by the responsible individual before official use.
- Should errors be found in AI-generated output, immediate correction or disposal is required, and users must be informed of such actions. Efforts must also be made to analyze the root cause of the problem and prevent recurrence.
- The potential for problems arising from AI utilization must be continuously monitored, and appropriate measures must be promptly taken if issues occur.
- When utilizing AI, it is important to acknowledge that legal and ethical responsibilities are inherent, and careful attention must be paid to prevent issues.
- Article 4 (Fact Verification and Validation)
- When utilizing AI for public services and content production, appropriate verification procedures must be undertaken.
- AI must not be used in a manner that distorts facts.
- Awareness of the possibility that AI-generated output may contain inaccurate or fabricated information is critical, and efforts must be made for fact verification and validation.
- Caution is required regarding the possibility of AI having been utilized in content acquired from or provided by third parties.
- For news and current affairs content, the principle is not to use AI-generated output for reporting and article writing without independent verification. However, AI may be used as an auxiliary tool to the extent that it does not distort the content's direction or facts, and does not mislead users.
- When utilizing AI for public services and content production, appropriate verification procedures must be undertaken.
- Article 5 (Transparency)
- When AI is utilized, users must be informed of this fact.
- Methods for indicating the utilization of AI shall comply with relevant laws and regulations.
- If AI-generated text, images, videos, or audio are utilized as core elements in public services and content production, the fact of AI utilization must be clearly communicated to viewers through appropriate means, and, if necessary, the source must also be indicated.
- Producers of public services or content utilizing AI are, in principle, required to preserve records of its use.
- When AI is utilized, users must be informed of this fact.
- Article 6 (Rights Protection)
- In the process of utilizing AI, care must be taken to ensure that the rights or interests of KBS and third parties are not infringed.
- Personal information protection laws must be complied with when utilizing AI.
- The 'KBS Information Security Guidelines' must be observed when utilizing AI.
- When inputting personal information into AI systems, consent from the data subject must be obtained, and its use must not exceed the scope of consent. Appropriate technical and administrative measures must be taken to prevent personal information leakage and unauthorized use as AI training data.
- Should content produced using AI be found to infringe upon the rights or interests of others, prompt measures such as deletion or correction must be taken.
- In the process of utilizing AI, care must be taken to ensure that the rights or interests of KBS and third parties are not infringed.
- Article 7 (Use of Portrait and Voice)
- In the process of utilizing AI, the portrait and voice rights of others must be respected.
- When utilizing the portrait, voice, or other personal elements of third parties in AI for public services or content production, relevant laws concerning their portrait and voice rights must be observed.
- Care must be taken to ensure that AI-generated output using the portrait or voice of others does not infringe upon their honor or personal rights.
- In the process of utilizing AI, the portrait and voice rights of others must be respected.
- Article 8 (Copyright Protection)
- When utilizing AI, care must be taken not to infringe upon the copyrights of others.
- If elements infringing on the copyrights of others are found during the use of AI in content production, prompt corrective action must be taken.
- Appropriate protection measures, such as the introduction of watermarks or copyright defense technologies, must be implemented to prevent KBS-owned content from being used without authorization as training data for third-party AI.
- When utilizing AI, care must be taken not to infringe upon the copyrights of others.