Generative Artificial Intelligence

Google teamed up with K-pop group Le Sserafim to launch a global campaign showing how its AI tool Gemini can help with everyday tasks. The music video-style ad blends pop culture and tech, featuring real-life uses of Gemini on Android phones.

A viral trend turning people into AI-generated action figures is capturing attention—but also raising alarms. Artists warn the tech threatens creative work, while explicit deepfakes continue to disrupt schools, politics, and public figures.

Kelly Yu, a Chinese-Canadian singer, has partnered with CreateAI to produce the world’s first AI-generated music video for her song "Werewolf," blending human creativity with AI technology. The project significantly reduced production time and costs, setting a new standard for future music video productions.

OpenAI is developing its own social media platform that integrates ChatGPT’s image generation with a social feed, aiming to compete with X and Meta. The project is still in early stages, with prototypes already being tested internally.

According to police sources cited by NHK and other outlets, the four individuals—ranging in age from their 20s to 50s—allegedly created and sold posters featuring indecent images of women through online auction platforms.

OpenAI's user base has skyrocketed to 800 million, driven by viral features like Ghibli-style image generation, with one million new users joining in just one hour. The company is now valued at $300 billion and is exploring ways to compensate artists for AI-generated content.

China has registered 346 generative AI services as of March 31, 2025, to regulate their development and ensure transparency. The new rules require AI providers to disclose model names and registration numbers, aiming to control AI’s impact on public opinion and social actions.

The music industry is increasingly concerned about the unauthorised use of its content to train generative AI models like Suno, Udio, and Mubert, raising issues of intellectual property rights and potential exploitation without consent or compensation.

The scammers employed a recurring strategy, initially targeting victims by posing as investment managers, only to later reappear as Europol agents promising to recover the stolen funds—for a fee.

DeepSeek's AI model R1 was found to give dangerous instructions like how to make malware and Molotov cocktails. Experts say the company focused too much on political censorship and not enough on real safety, raising global security concerns.