BBC의 AI 저널리즘…“신뢰도 리스크 관리·교육 체계가 먼저”

A recent report by the British Broadcasting Corporation (BBC) highlights the broadcaster's approach to managing the risks and building confidence in the use of artificial intelligence (AI) in journalism. The Korea Media Development Institute (KMDI) released the report titled "BBC's AI Governance and Journalism Innovation: Designing Trust in the Age of Generative AI" on May 25th.

Contrary to the competitive atmosphere in the media industry, the BBC prioritizes the establishment of principles and education systems over technology adoption. At the beginning of their AI adoption journey, the broadcaster strictly limited the use of AI by front-line reporters and production staff. Only after establishing organizational guidelines in February 2024 through their "Guidelines for AI Use" did they allow its utilization.

The BBC's cautious step-by-step approach aims to integrate AI into news and broadcasting production systems while upholding public broadcasting principles. This approach stems from concerns about the potential impact of AI on journalistic credibility. A self-conducted BBC study from last year revealed that answers generated by four popular generative AI services, including ChatGPT, Copilot, Gemini, and Perplexity, contained significant issues (51%) such as factual inaccuracies, incorrect numbers, and false date references. Additionally, 13% of the answers cited non-existent news sources.

To address these concerns, the BBC has outlined a set of core principles that form the foundation of their "AI Governance" framework: the public interest, the expertise and creativity of journalists, transparency for audiences, and openness. This framework supports the development of editorial guidelines, transparency standards, risk management procedures, and comprehensive education and support systems.

The evaluation also includes a dedicated team of experts who analyze the risks associated with AI use across various domains, such as media ethics, privacy, and legislation. Furthermore, the BBC provides in-house training programs on AI usage and practical editing judgment, ensuring that journalists are responsible for the final outcomes. For example, the BBC's audio streaming service uses AI to create subtitles for viewers with hearing impairments, as well as edit content from local news ecosystems into BBC's style.

The report concludes that the critical aspect lies not in what AI produces, but how its results are validated and approved within the broadcaster's editorial judgment. In the era of AI, the key to maintaining journalistic competitiveness is ensuring that the technology is responsibly utilized while preserving trust.