CNN joined a new deal allowing Google to use its content to train AI models. This partnership follows similar moves by the Financial Times and Axios. Google said over 400 international publications now participate in the programme. This marks CNN’s first such agreement since Paramount Skydance acquired Warner Bros Discovery.
CNN partners with Google AI
News publishers have been trying to cope with falling referral-traffic from Google for a good few years now. The search engine giant prioritizes having Gemini answer the queries over linking to websites with that information, and it is an uphill battle for publications to stop AI companies from using copyrighted content to train their models. While lawsuits against Google, OpenAI and Claude—to name just three companies—have piled out over the years, other publications are taking a path of no resistance. This involves allowing Google to use their content to train AI models, in return for which Google pays them for participating. The most recent major news publishers to join this are CNN, the Financial Times, and Axios, among others. For CNN, this is the first deal the company has entered into since Paramount Skydance acquired Warner Bros Discovery. According to Google, over 400 international publications are now a part of the programme.
G7 nations guard against AI attacks
In light of the damaging attacks on businesses carried out by rogue AI agents in recent months, finance leaders of the G7 nations have decided to find a mechanism to guard against cyber security threats. “The G7 finance deputies have been discussing the idea of doing tabletop exercises on frontier AI, which was signed off at the ministerial level,” said the Japan finance minister Satsuki Katayama. She was attending a meeting of central bank governors and finance ministers of the G7 countries on the sidelines of an International Monetary Fund (IMF) gathering in Bangkok, Thailand. The exercise will try and evaluate the robustness of current protocols regarding cyber safety of financial systems, and revise them if required.
Demystifying the AI ‘Black Box’
The AI industry is no stranger to long essays, written by bosses such as Sam Altman and Mark Zuckerberg, as well as researchers working in their companies. While most of these have been in the form of free-flowing genuflections on the nature, promise and dangers of AI, few have caused as much of a flutter as a 26,000 word essay published a few days ago by the researcher Nick Cammarata. The essay, titled Neural Networks are Already Clean, is Cammarata’s attempt to understand and define the true nature of current AI. His claims that generative AI models are not impenetrable “black box” systems and can be understood and controlled, flies in the face of all current assumptions.