[The Editors' Verdict] Reporters' Legwork Reduced to Free Data

원본보기 아이콘

Last weekend, I went backpacking to Guleopdo, an island off Incheon often called “the Galapagos of Korea.” It rained from early morning and the wind was quite strong, but fortunately, the ferry was not canceled. Upon arrival, I shouldered a 13kg backpack and hiked for about an hour along a mountain trail before reaching my campsite. The unexpectedly steep slopes, rugged rocky paths, and rain-lashed winds felt overwhelming before I unpacked. Yet, sitting in my tent, gazing at the sea with a can of beer in hand-that’s exactly why I willingly choose backpacking over the comfort of a hotel.

The scene of a reporter’s work is much like this strenuous mountain trek with a heavy backpack. Reporting is inherently an endless series of uncertainties. Dozens of phone calls to sources often end with cold rejections. Even data obtained after much difficulty can lose its news value if someone else has already used it. Only after a grueling process of doubting, trimming, and cross-verifying fragments of information collected through reporting does a single piece of fact finally emerge. The news we encounter every day is not just a seamless string of text, but a hard-won trophy earned through a battle with uncertainty.

However, these precious assets are being stolen all too easily by the artificial intelligence (AI) of global big tech companies powered by immense capital. AI snatches only the summit photos-taken as media organizations bear the weight of their backpacks-while thoroughly erasing the grueling investigative path beneath them. Decades of news content, built up through the sweat and effort of media outlets, is being treated as free data under the pretext of “AI advancement.” According to a 2023 white paper published by the News Media Alliance (NMA), which represents over 2,000 newspapers in the United States and Canada, large language models (LLMs) such as ChatGPT used news content at rates five to one hundred times greater than ordinary web content. While many news organizations, having realized the gravity of the recent situation, have started blocking crawlers, AI’s unauthorized training continues to bypass such measures and is still ongoing.

The answers generated by AI, which has grown smarter by consuming news for free, carry neither the scent of sweat nor a hint of hesitation. In articles written by journalists, you’ll often find expressions like “the circumstances suggest this is likely” or “we cannot rule out the possibility.” These phrases are sometimes misunderstood as the reporter’s “speculation,” but they are, in fact, devices to signal to readers that the information has been logically connected through careful reasoning. In AI-generated responses, one cannot detect this “imperfect but genuine effort to convey the truth.” Even the byline is erased, and fatal errors are hidden behind technical jargon such as “hallucination.”

Global news organizations are joining forces to prevent the indiscriminate theft of news content by AI. In 2023, The New York Times filed a copyright lawsuit against OpenAI and Microsoft, and in 2024, The Wall Street Journal and New York Post sued Perplexity. In February of this year, leading British media outlets including the BBC, Financial Times, Guardian, Daily Telegraph, and Sky News formed a joint group called SPUR (Standards for Publisher Usage Rights). Even rival outlets with completely opposite political leanings and target audiences, such as the progressive Guardian and the conservative Telegraph, have united. The sense of crisis to prevent the collapse of journalism has transcended ideological differences and gone beyond the mere survival of news organizations.

In contrast to this trend, Korean news organizations’ response has been lukewarm and fragmented. The only copyright lawsuits have been filed by the three major terrestrial broadcasters-KBS, MBC, and SBS-against Naver in January last year and against OpenAI in February this year. Meanwhile, some broadcasters have chosen to sign AI partnerships with Naver, each seeking their own path to survival.

Currently, global AI companies are slowly turning up the heat with news organizations in the pot. Korean news outlets seem to have lost even the will to fight back, as they witness data being used for unauthorized training and their web traffic being siphoned off before their eyes. Although Korean media should unite to confront global AI companies, all attention has shifted instead to the evaluation system of the major portal, which will be fully implemented from May. This is a result of being dependent on the traffic and syndication fees provided by the portal. Once the water reaches its boiling point, journalism may end up being reduced to anonymous data with no one’s name attached. By then, even the spirit of journalism-like the romance of backpacking-will have evaporated without a trace, just like the water in a boiling pot.

© The Asia Business Daily(www.asiae.co.kr). All rights reserved.