AI Scraping Wikipedia for Free Is 'LuckyWiki'... Angry Wikipedia Says "Pay Up"
Free Online Encyclopedia 'Wikipedia'
Requests Paid Service Use from AI Developers
"Human Traffic Down, Bot Access Up"
Wikipedia, the free online encyclopedia, has requested that artificial intelligence (AI) developers use its paid services. According to Yonhap News on November 10 (local time), citing the U.S. IT media outlet TechCrunch, the Wikimedia Foundation, which operates Wikipedia, has asked AI developers to use its paid product, the "Wikimedia Enterprise Platform."
According to the foundation, AI developers have been scraping large amounts of Wikipedia content to train generative AI models based on large language models (LLMs). To build high-performing AI models, it is essential to train them with high-quality data. Wikipedia’s content is considered to meet the criteria for training data because it is vast in volume, objective, and reliable.
The foundation stated, "Visits from human users have decreased by 8% compared to last year, while visits estimated to be from bots are increasing." It added, "There have also been attempts by bots to disguise themselves as humans and evade 'bot detection.'" The foundation urged, "If developers choose the paid product, they can access content at scale without putting a severe burden on Wikipedia’s servers."
Additionally, the foundation requested that AI platforms clearly cite Wikipedia as a source when providing answers that reference its content. The foundation explained, "For people to trust the information shared on the internet, platforms must clearly disclose the sources of information and increase opportunities for users to visit those sources." It also warned, "If visits from human users to Wikipedia decrease, there will be fewer volunteers to grow the content and fewer individual donors to support it."
Recently, the U.S. online marketing platform Semrush released an analysis of which websites were cited in over 150,000 answers generated by AI models such as Google’s AI Mode, AI Overview, ChatGPT, and Perplexity up to June this year. Wikipedia ranked second, cited in 26.3% of cases, following Reddit, the largest online community in the United States.
Hot Picks Today
"Samsung Electronics Employee With 100 Million Won Salary" to Receive 600 Million Won Bonus... Estimated Tax Released
- At President Lee's Call to "Give Enough to Shock," Whistleblower Rewards Become a Real Lottery
- Lived as Family for Over 30 Years... Daughter-in-Law Cast Aside After Husband's Death
- "Is This Car Waterproof?" US Man in His 70s Makes Dramatic Escape After Driving Cybertruck Into Lake, Trusting Its Features
- Appearing in a Leather Jacket, Jensen Huang Hastily Eats $6 Noodles on the Street... Shop Instantly Becomes a Hotspot
There is also a growing number of media companies filing lawsuits against corporations, claiming that AI summaries have reduced their traffic and revenue. In September, Penske Media, which owns Rolling Stone and The Hollywood Reporter, filed an antitrust lawsuit against Google in the U.S. District Court for the District of Columbia. They argued that Google’s AI Overview, which appears at the top of search results, diverts traffic away from media outlets and reduces their revenue. In contrast, Google countered that AI Overview increases search efficiency and drives traffic to a wider variety of sites.
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
