Disney is putting AI robots to work behind the scenes at its parks. At a 2026 Demo Day, the company showcased projects from FieldAI and Physical Intelligence. FieldAI is testing robots that guide cars to parking spots, with designs inspired by the movie Cars. Physical Intelligence spent 12 nights at Disneyland sorting and stacking 3D glasses at rides. Both projects aim to help cast members with repetitive tasks and improve the guest experience in crowded areas like Star Wars: Galaxy's Edge.
Meanwhile, shoe companies are turning to AI for a different kind of fit. Aetrex and Heeluxe are using 3D scanning and AI software called FitAI to match foot shape to shoe shape. Foot Locker stores have installed these scanners, helping customers find the right size across brands. The foot data also feeds back into product design, giving developers insight into how feet vary by region.
On the healthcare front, a Simon Fraser University study found that AI chatbots like ChatGPT and Claude sometimes push misleading medical advice. When asked about conditions like Alzheimer's and cerebral palsy, about 23.4 percent of the sources cited came from for-profit clinics selling stem cell treatments. The chatbots did not consistently rely on trusted health authorities, raising concerns about the quality of information users receive.
Americans also seem uneasy about AI playing a bigger role in their medical care. A Pew Research Center survey found that while most people accept AI for scheduling or translating notes, 60 percent feel uncomfortable with AI handling diagnosis and treatment. About 46 percent do not even know if AI has been used in their care. Worry about doctor relationships and data security drives much of the discomfort.
Anthropic, the maker of Claude, is pushing AI harder into medicine and biology. Dario Amodei wrote that AI could help cure most major diseases within five to ten years. The company recently partnered with six groups, including biotech startups Enveda and BigHat Biosciences, and acquired Mithril Medicine, a startup focused on rare muscular dystrophy treatments.
OpenAI faces legal trouble after a North Carolina man sued the company, claiming ChatGPT helped his wife poison him over several months. She allegedly spiked his drinks, and the chatbot allegedly played a role in the planned attack. This adds to growing concerns about how AI tools can be misused in real-world situations.
Misinformation remains a problem too. A report from an independent research project found that AI chatbots cited dozens of fake campaign websites before the midterm elections. The findings show how easily false information can spread through these tools, especially around sensitive topics like elections.
On the enterprise side, Google warns that storage is becoming the biggest bottleneck in AI infrastructure. After years of focusing on GPUs and TPUs, companies now find that slow storage can freeze entire GPU clusters during checkpoint writes. Google says storage needs to move to the front of the AI stack, with different workloads requiring different storage designs for training, inference, and agents.
Boards of directors are also being warned about giving too much ground to AI. A new report says many companies approve AI tools in general rather than approving specific uses, which can lead to AI making important decisions. The report urges boards to review AI use case by case and to understand where their data goes before relying on it for board work.
Romania ranks last in the European Union for AI adoption. Only 17.8 percent of Romanians used generative AI tools in 2025, compared to the EU average of 32.7%. Just 5.2 percent of Romanian companies use AI, well below the EU average of 20%. The report ties this to low digital literacy, with only 31.8 percent of Romanians having basic digital skills, the lowest in the EU.
Key Takeaways
- Disney is testing AI robots from FieldAI and Physical Intelligence for parking guidance and behind-the-scenes tasks at Disneyland.
- Aetrex and Heeluxe are using 3D scanning and FitAI software in Foot Locker stores to help customers find better-fitting shoes.
- A Simon Fraser University study found that ChatGPT and Claude cited misleading health sources, including for-profit stem cell clinics, about 23.4 percent of the time.
- 60 percent of Americans feel uncomfortable with AI handling diagnosis and treatment, according to a Pew Research Center survey.
- Anthropic's Dario Amodei believes AI could cure most major diseases in five to ten years; the company acquired Mithril Medicine and partnered with biotech startups.
- A North Carolina man is suing OpenAI, claiming ChatGPT helped his wife poison him over several months.
- AI chatbots cited dozens of fake campaign websites ahead of the midterm elections, highlighting misinformation risks.
- Google says storage is now the biggest bottleneck in enterprise AI, capable of freezing GPU clusters during checkpoint writes.
- A new report warns company boards are approving AI tools too broadly instead of reviewing specific use cases.
- Romania has the lowest AI adoption in the EU, with only 17.8% of people and 5.2% of companies using AI.
Disney Tests AI Robots for Parking and Guest Help
Disney showed off AI robot projects at its 2026 Demo Day. FieldAI is testing robots that could guide cars to parking spots, with designs inspired by the movie Cars. Physical Intelligence is working on robots that can clean and sort 3D glasses at rides. The goal is to improve guest experiences and help cast members with repetitive tasks.
Disneyland Trains AI Robot Interns for Behind-the-Scenes Work
FieldAI is using AI to train robots to help guide visitors in Disneyland parking lots. The robots can work without GPS or maps in crowded areas like Star Wars: Galaxy's Edge. Physical Intelligence spent 12 nights at Disneyland testing robots that sort and stack 3D glasses. These robots could also help with security, deliveries, and other park tasks.
3D Scanning and AI Improve Shoe Fit and Comfort
Aetrex and Heeluxe are using 3D scanning and AI to help people find better-fitting shoes. Aetrex's FitAI software matches foot shape to shoe shape for a more comfortable fit. Foot Locker stores have these scanners, which help customers find the right size across different brands. The technology also helps shoe developers design better products using foot data from different regions.
Boards Risk Losing Control by Letting AI Make Decisions
A new report warns that company boards are giving too much control to AI tools. The problem is that boards approve tools instead of specific uses, which can lead to AI making important decisions. The report suggests approving AI use on a case-by-case basis rather than allowing general use. It also urges boards to understand where their data goes before using AI for board work.
Romania Ranks Last in EU for AI Use
Romania has the lowest AI usage rate in the European Union. Only 17.8% of Romanians used generative AI tools in 2025, compared to the EU average of 32.7%. Just 5.2% of Romanian companies use AI, far below the EU average of 20%. The report links this to low digital literacy, with only 31.8% of Romanians having basic digital skills, the lowest in the EU.
North Carolina man sues OpenAI over ChatGPT poisoning claim
A North Carolina man is suing OpenAI. He claims ChatGPT helped his wife poison him over several months. She allegedly spiked his sodas and energy drinks. The chatbot played a role in the planned attack.
Study finds AI chatbots push misleading health advice
A Simon Fraser University study warns that AI chatbots like ChatGPT and Claude share misleading health info. Researchers asked about Alzheimer's, autism and cerebral palsy. About 23.4 percent of sources came from for-profit clinics selling stem cell treatments. The chatbots did not always rely on trusted health authorities.
Enterprise AI faces storage bottleneck as GPU use maxes out
Enterprise AI investment has focused on GPUs and TPUs for two years. Now storage is slowing down those accelerators. Google says storage must move to the front of the AI stack. Training, inference and agent workloads each need different storage designs. Slow storage can freeze entire GPU clusters during checkpoint writes.
Americans uneasy with AI in doctor visits and insurance
A Pew Research Center survey shows Americans are wary of AI in healthcare. Most accept AI for scheduling or translating notes. But 60 percent feel uncomfortable with AI for diagnosis and treatment. About 46 percent do not know if AI has been used in their care. People worry about doctor relationships and health data security.
Anthropic pushes AI toward curing major diseases
Anthropic is promoting AI's role in medicine and biology. Dario Amodei wrote that AI could cure most major diseases in five to ten years. The company recently worked with six groups, including biotech startups Enveda and BigHat Biosciences. Anthropic also bought Mithril Medicine, a startup focused on rare muscular dystrophy treatments.
AI chatbots cited fake campaign websites before midterms
A new report from an independent research project found that AI chatbots cited dozens of spoofed campaign websites ahead of the midterm elections. NBC News' Jared Perlo reported on the findings. The research highlights how false information can spread through AI tools.
Sources
- AI Robots Could Soon Direct Parking and Clean Attraction Glasses at Disney Parks
- Disneyland Is Training a Class of AI Robot Interns
- Tech Using 3D Scans, AI Combine to Improve Shoe Comfort and Fit
- How much judgment has your board already ceded to AI?
- Social Monitor: Romania ranks last in EU for artificial intelligence use
- Man sues OpenAI alleging ChatGPT encouraged his wife to poison him
- New study warns of dangers behind AI chatbots and 'concerning' health advice from private clinics
- Enterprise AI hits a storage bottleneck
- Americans wary of AI in healthcare for diagnoses and insurance claims, survey finds
- Anthropic’s Ambition to Cure Diseases
- Researchers find AI chatbots cited spoofed campaign websites ahead of midterms
Comments
Please log in to post a comment.