Image Generation

Today's Stories

1. 1

   Dataland Opens World's First AI Art Museum in Los Angeles

   Artist Refik Anadol and partner Efsun Erkılıç opened Dataland, billed as the world's first "museum of AI arts," in downtown Los Angeles on June 20. The opening exhibit, Machine Dreams: Rainforest, welcomed more than 10,000 visitors in the first two weeks. Dataland challenges the perception that AI art is purely low-quality generated content. The gallery demonstrates a methodical approach to AI art creation—Anadol's team trained their own AI models using 5 petabytes of raw data collected with researcher consent from the Amazon and other rainforests, avoiding the unlicensed data extraction that has drawn lawsuits against major tech firms.

   The exhibit uses wearable biosensors to make artworks responsive to visitors' movements, heart rate, skin temperature, and other biometric data. The gallery commits to deleting visitor data upon departure, treating personal information as a form of memory to be respected rather than harvested for surveillance.

2. 2

   Over 70% of Japan accommodations report staff shortages amid tourism surge

   A government white paper on tourism released Friday found that 72.2% of 522 accommodation facilities surveyed between December and January reported labor shortages. Midsize facilities with annual sales between ¥100 million and ¥1 billion were hardest hit, with 77.1% experiencing understaffing. As Japan pursues its goal of 60 million inbound tourists annually by 2030—up from 42.68 million in 2025—chronic understaffing threatens the sustainability of tourism as a strategic industry. Understaffed properties report increased workload on existing workers (79.3%), higher hiring costs (50.4%), and forced service cuts (40.6%), which may degrade the visitor experience and competitiveness.

   The white paper identifies low pay and insufficient days off as root causes, and recommends fundamental fixes: raising salaries, improving benefits, creating women-friendly workplaces, and investing in digital tools. Temporary measures like hiring foreign nationals and part-time workers have provided only short-term relief.

3. 3

   Midjourney demands studios reveal their own AI training to prove fair use

   Midjourney has filed a motion to review a mid-June court ruling that blocked the image-generation AI company from requiring Disney, Universal, and Warner Bros. to disclose how they use AI themselves. The studios sued Midjourney in 2025, alleging the company infringed copyrights of franchises including Superman, Scooby-Doo, and Yoda by training its models on publicly available images. Midjourney argues that if the studios are themselves training AI models on third-party copyrighted works scraped from the internet, that conduct would serve as evidence the industry considers such use to be fair use under copyright law. The outcome of this case may determine how copyright law applies to AI training across the industry, with consequences for artists, art markets, and companies building AI systems.

   US copyright law allows unlicensed use of copyrighted material under a "fair use" exception based on several factors—purpose, nature of the work, how much was copied, and market impact. Courts have ruled in some prior AI cases that training is transformative fair use, but fair use remains an unsettled area of law; it is unclear when the California district court will rule on Midjourney's motion.

4. 4

   Meta releases Muse Image, ranks #2 in AI photo generation

   Meta launched Muse Image, its first image generation model from Superintelligence Labs, which works as an AI agent that iteratively refines outputs by calling tools like code generation and web search. On the Image Arena evaluation platform, Muse Image ranks second in human preference scores for text-to-image and for both single- and multi-image editing, behind only OpenAI's GPT Image 2. The model is now available in Meta AI app, on meta.ai, in Instagram Stories in the US, and in WhatsApp—reaching Meta's largest user bases. However, a newly introduced feature allows users to @-mention public Instagram accounts in prompts so Meta AI pulls photos from those profiles to generate images of that person with no consent required. The feature is on by default and appears set to face regulatory scrutiny in Europe under GDPR and the EU AI Act's deepfake labeling rules, which take effect August 2, 2026.

   Meta's invisible watermark system, Content Seal, survives cropping and compression, but whether a machine-readable watermark alone satisfies the EU AI Act's requirement that AI-generated images resembling real people be labeled in a way recognizable to affected people remains an open question. Images already generated will not be deleted even if users opt out of the feature.