Taylor Swift suspected of using AI in new album promotion campaign
Taylor Swift just released her 12th album, The Life of a Showgirl, with a Google scavenger hunt. Fans had to find 12 mysterious doors in cities, scanning QR codes to unlock videos containing hints. However, many fans said that these videos looked like they were created with artificial intelligence (AI), leaving them skeptical and dissatisfied.

Taylor Swift in bright stage lights, amid suspicions of using AI for the promotional campaign for her album "The Life of a Showgirl". (Source: Getty Images)
Google has not confirmed whether it used AI in this campaign, but Swift and Google have collaborated before. The controversy has increased since Swift warned about the risk of spreading misinformation through AI, especially after she was accused of using AI to manipulate images of her supporting President Donald Trump last year.
OpenAI partners with AMD to boost AI infrastructure
On October 6, OpenAI announced a strategic partnership with chipmaker AMD to build a next-generation artificial intelligence infrastructure. Under the agreement, AMD will provide the high-performance Instinct MI450 graphics chip line, expected to launch next year, with a total capacity of up to 6 gigawatts. The first batch of chips will be deployed in the second half of 2026.
Notably, OpenAI has the option to buy up to 160 million shares of AMD stock—equivalent to about 10% of the company’s stock—if certain capacity and stock price milestones are met. CEO Sam Altman called it “a huge step forward in realizing the potential of AI.”
The deal not only helps AMD compete with Nvidia, but also shows the huge demand for computing power in the AI era.
Tencent surpasses Google with Hunyuan Image 3.0 image-generating AI model
Tencent has just launched Hunyuan Image 3.0 - the world's largest image-generating AI model with 80 billion parameters - and quickly took the No. 1 spot on the LMArena rankings, surpassing its rival Nano Banana from Google DeepMind.

Image of a Star Ferry-inspired spacecraft traveling through a wormhole, created by Google's Nano Banana. (Source: Google)
Trained on 5 billion image-text pairs and 6 trillion tokens, Hunyuan Image 3.0 currently only supports image generation from text but will soon add image editing and transformation capabilities. The model is open-sourced on GitHub and Hugging Face, and costs $0.10 per megapixel—higher than competitors like ByteDance’s Seedream 4.0.
Tencent claims the tool will help creators cut their work time from hours to just minutes.
Source: https://vtcnews.vn/cong-nghe-7-10-taylor-swift-bi-nghi-dung-ai-trong-chien-dich-album-moi-ar969661.html
Comment (0)