Generative AI, one of them ChatGPT, was in several speaker’s presentation at the opening day of the 74th World News Media Congress in Taipei, Taiwan, on June 28, reflecting how the technology’s use has “exponentially increased,” and its effects in the newsrooms more apparent. The interesting plenary session on “Generative AI: Media savior or existential threat?”  was presented by Ezra Eeman, director of Strategy and Innovation, NPO, Netherlands, Laura Ellis, head of Technology Forecasting, BBC, UK, Tav Klitgaard, CEO Zetland, Denmark, Agnes Stenborm, head of IN/LAB Schibsted, Sweden, and Pierre Louette, president and CEO, Les Echos-Le Parisien, France. Leading the conversation on AI was the summary of the session: “Generative AI (for example, ChatGPT) has already sparked a revolution in the creative industries – and has the potential to disrupt media businesses, journalism and the information ecosystem. What are the upsides for news publishers and journalism, and where are commercial and ethical safeguards needed?” Cited as one of the defects of generative AI was “hallucinations” with an example showing how it created content that included false information on a topic.  Thus, the need for a human to check on what the AI has created. Another example showed that generative AI has bias, showing a photos of inmates who had black and brown skin tones, in reply to a query on images related to crime. “Generative AI is not reliable and introduces new points of friction and failure, such as hallucinations, lack of critical judgement and copyright infringement.” The other effects of AI were also presented.  These are: It has the potential to level the professional playing field and creative skills become “superdistributed.” Second, it increases the speed to gather and package news, thus increasing efficiency through faster news cycles. One of the more serious effects of AI was cited as its “generating exponential noise.”  An example shown was a story reporting that “Dozens of news sites are using AI to churn out questionable content,” and some of them “publishing hundreds of articles a day.” Talk about AI started as soon as the opening ceremonies hours before the afternoon plenary session on the topic. Felrnando de Yarza Lopez Madrazo, president of WAN IFRA, said: “We must recognize the immense possibilities, and at the same time, huge challenges and understandable fears, of Artificial Intelligence and what it holds for our collective futures. What AI means for our newsrooms is not yet clear, but while it promises to be a game changer we must be incredibly wary about not repeating mistakes of the past. A rush to embrace new technology without considering the implications, especially concerning AI, is potentially catastrophic.” WAN-IFRA is the global organization of the world's press, comprising 3,000 news publishers and technology companies and 60 national publishers' associations representing 18,000 publications in 120 countries. In the afternoon session of the first day of the congress, a one-on-one with Joseph Kahn, New York Times executive editor, also touched on AI in the newsroom. He said the use of AI can be a burden as someone will be needed to fact check the details in a published story which AI had produced. The World Media Congress which gathered 950 participants from 58 countries, is held to deliberate on the future of news media in the changing world. It will end on June 30. This is the first time this is held in Taiwan after a two-year delay caused by the pandemic.