Updated April 25, 2026 at 8:26 PM EDT
Los Angeles has been the nexus of splashy museum developments lately, including this week's opening of the imposing David Geffen Galleries at the Los Angeles County Museum of Art and the June 20 unveiling of Dataland. There is also the upcoming launch of the Lucas Museum of Narrative in September.
Billed by its co-founder, Turkish-American artist Refik Anadol, as "the world's first museum of AI arts," Dataland will be located at The Grand LA, the Frank Gehry-designed complex in downtown Los Angeles' Grand Avenue Cultural District.
The high-profile artist is best known for creating an epic generative AI piece for New York's Museum of Modern Art in 2022, which used AI to interpret 200 years of the museum's collection. Anadol is also recognized for projecting a massive installation based on the Los Angeles Philharmonic's archives onto the exterior of the Walt Disney Concert Hall in Los Angeles.
"Building the world's first Museum of AI Arts has been the journey of a lifetime for me," said Anadol in a post on Instagram. "We cannot wait to welcome you all to step inside these immersive, nature-inspired data worlds and experience the true creative potential of human and machine collaboration."
The privately-funded museum spans 35,000 square feet. Its five galleries have been explicitly designed for fully immersive, 360-degree art experiences.
Anadol's Machine Dreams: Rainforest, the museum's inaugural exhibition, is an immersive, audiovisual experience based on millions of images and sounds of nature. It's "a narrative of a deepening relationship between machine intelligence and the natural world," according to the museum. The installation was inspired by a visit Anadol and museum cofounder Efsun Erkılıç made to the Amazon rainforest.
Arriving on the scene at a time of pushback
While Dataland may be the first major museum devoted exclusively to AI art, institutions like the Gray Area Foundation in San Francisco place a lot of emphasis on the medium, and major museums such as ZKM Center for Art and Media in Karlsruhe, New York's Museum of Modern Art and the Design Museum in London have also prominently featured AI art.
Barry Threw, Gray Area executive and artistic director, told NPR that Dataland is compelling because it renders "complex data as experience."
"At a time when so much of our world is shaped by technology, art can help us make sense of it," Threw said. "Anadol's work, especially with Dataland, raises a question Gray Area often returns to: can digital art allow us to look beyond aesthetics and think more deeply about the systems behind it?"
The arrival of this new AI-centric art museum comes at a time when the medium continues to provoke widespread criticism for its lack of true human agency.
"Let's build a museum based on instructions people give to AI and call it art. It's not and it never will be," commented artist Thomas Brummett, who uses digital techniques as part of his process, on Instagram. "At best, it's just second rate entertainment." Brummett has pieces in collections at the Museum of Fine Arts in Houston, the Philadelphia Museum of Art, and the Museu de Arte Moderna in Rio de Janeiro.
AI art also regularly elicits pushback for its environmental impact, the unethical training of its large language models on human intellectual property, and the inherent biases of its output.
"AI art often uses models that lack safe content with diverse representation, increasing gender and racial biases," said Boston-based artist Nettrice Gaskins, who is best known for her complex and colorful AI-driven images, in an email to NPR. "The process is far from perfect. AI hallucinations, which are not faithful representations of reality, often contain imagined, stereotypical, or unrealistic elements."
Gaskins also said Dataland may help "shift the discussion about the value of AI art, which would benefit more artists who use some aspect of AI in their work."
Responding to concerns about copyright and biased content
On the issue of data and copyright, Anadol told NPR that the dataset that underpins Machine Dreams: Rainforest – what the founders call a "Large Nature Model (LNM)" – is trained on data collected first-hand from 16 rainforests around the globe. It is also trained using data partnerships with the Smithsonian, the Cornell Lab of Ornithology, Getty, iNaturalist, and London's Natural History Museum.
"The best way to achieve responsible curation is to build our own models and be radically transparent about where our data comes from," Anadol said in an email.
Asked about how Dataland will handle environmental custodianship, Anadol said, "Sustainability is not a constraint we work around; it is a condition we build from."
"The LNM runs entirely on a specialized Google Cloud server in a low-CO2 compute zone in Oregon, utilizing 87% carbon-free renewable energy. The energy required to generate a visitor's entire stay is roughly equal to charging a smartphone," Anadol also said.
Regarding concerns about AI's ability to "hallucinate" and churn out harmful or biased content, Anadol said a machine learns from what it is trained on.
"So the question of what we teach it—and whose knowledge we draw from—is everything," he said.
Anadol added that his LNM is built using permission-based datasets from partner organizations, as well through a collaboration with the Yawanawá people of the Amazon.
"Their ancestral understanding of the forest shaped our entire approach to what it means to truly learn from nature with respect and care," he said.
Copyright 2026 NPR
