Utah is moving forward with AI in health care through three new clinical pilots. The state partnered with Intermountain Health, University of Utah Health and outside evaluators under its Pro-Human AI Initiative. The pilots will test AI tools for acne telehealth, pelvic health physical therapy and prescription renewals over 12 months with state oversight. Master agreements on data privacy and AI governance were also signed, as leaders note AI could help address a shortage of health care providers by 2035.
OpenAI is set to appear before Australia's Joint Select Committee on Artificial Intelligence for four days of hearings. The company must explain how it prevents AI agents from accessing non-public Australian data. Senator David Pocock and chair Jo Briskey also plan to press OpenAI on its response to the Services Australia incident, highlighting growing regulatory scrutiny of how AI firms handle sensitive government data.
Google Research published a report warning that personal AI agents can leak sensitive data without any external hacking. The study, co-authored with Cornell University and UC Berkeley, found that services like Meta's Muse and OpenAI's Dots store user information that can persist even after deletion, referred to as ghost information. Researchers recommend limiting AI permissions and requiring user confirmation before sharing sensitive data.
Financial firms still struggle to measure the return on investment for AI. A Sibos poll found 61% of companies have little visibility into AI token usage despite tracking overall spending, and only 13% actively manage costs at the token level. HSBC and Sumitomo Mitsui Banking Corp. shared examples of AI deployments that improved efficiency and reduced costs, but unclear resource tracking remains a widespread problem.
Companies also face challenges when AI is cited as a reason for layoffs. A Financial Times examination found that firms often mix AI justifications with cost-cutting and restructuring. HR should treat claims that AI causes job losses as testable ideas rather than settled facts. Deloitte's 2026 survey found only 6% of leaders said their organizations are improving human-AI work design, even though 85% called workforce adaptability critical. Gallup's August 2026 research found 99% of CHROs see AI as important to strategy, but 50% lack confidence in managers guiding employees' AI use.
Key Takeaways
- Utah launched three clinical AI pilots with Intermountain Health and University of Utah Health, each running 12 months under state oversight
- OpenAI must appear before Australia's Joint Select Committee to explain how it prevents AI agents from accessing non-public data
- Google Research found that Meta's Muse and OpenAI's Dots can leak sensitive data through persistent ghost information even after deletion
- 61% of financial firms have little visibility into AI token usage, and only 13% actively manage token-level costs
- Only 6% of leaders say their organizations are improving human-AI work design despite 85% calling workforce adaptability critical
- 99% of CHROs view AI as important to strategy, but 50% lack confidence in managers guiding employees' AI use
- Utah's AI health care pilots target acne telehealth, pelvic health physical therapy and prescription renewals
- Applied Digital is expanding to Finland with up to 1 GW of potential power capacity for a new AI campus
- Kevin Scharp estimates an 89% chance of human extinction within 12 to 18 months if AI continues on its current path
- 86% of college students use AI regularly for coursework, prompting calls for stricter AI guidelines in higher education
Utah launches health care AI pilots with outside evaluators
Utah announced three new clinical AI pilots as part of its Pro-Human AI Initiative. The state partnered with Intermountain Health, University of Utah Health and outside evaluators to test AI tools safely. Pilots include acne telehealth, pelvic health physical therapy and prescription renewals. Each pilot will run for 12 months with state oversight.
Utah expands AI in health care through new agreements
Utah's Office of Artificial Intelligence Policy signed master agreements with Intermountain Health and University of Utah Health. The deals cover data privacy and AI governance for testing new tools. Three pilot programs were also approved, including AI for acne, postpartum physical therapy and prescription refills. Leaders say AI could help address a shortage of health care providers by 2035.
OpenAI must explain how it stops AI hacking Australian data
Australia's Joint Select Committee on Artificial Intelligence will hold four days of hearings. OpenAI's Jason Kwon, Adam Cohen and Peter Anstee will appear. The committee wants OpenAI to explain how it will prevent AI agents from accessing non-public Australian data. Senator David Pocock and chair Jo Briskey also pressed the company on its response to the Services Australia incident.
Optogenetics won a Nobel, not a business model
Karl Deisseroth, Peter Hegemann and Georg Nagel won the 2026 Nobel Prize in Medicine for optogenetics. The award validates the science of controlling neurons with light. However, the article says startups still face major barriers. These include viral delivery, light penetration in tissue and limited human data. Turning the research into products remains far from proven.
Philosopher Kevin Scharp on AI and America's future
Kevin Scharp, a philosophy professor at the University of Illinois Urbana-Champaign, spoke about AI risks as part of America's 250th birthday reflections. He says superintelligence could pose an existential threat. He estimates an 89% chance of human extinction within 12 to 18 months if AI continues on its current path. He also notes AI is now solving problems that stumped human mathematicians.
Applied Digital Expands to Finland with 1 GW AI Campus
Applied Digital announced its first international expansion into Finland, securing up to 1 gigawatt of potential power capacity. The company plans to build a large-scale AI campus using Finland's favorable climate and energy ecosystem. The project will follow a disciplined investment model based on customer demand and regulatory conditions. This move opens a pathway for future expansion in European AI markets.
Ball State Daily Calls for Stricter AI Guidelines in Colleges
An opinion piece in the Ball State Daily argues that colleges need stricter rules on artificial intelligence use. The author points out that nearly 86% of college students use AI regularly for coursework. The article warns that AI use weakens critical thinking, communication skills, and job preparation. It calls on schools like Ball State to update their AI policies to better protect student learning.
Martha's Vineyard Tourism Breakfast Explores AI Strategies
Leaders in Martha's Vineyard tourism discussed AI strategies at the State of the Island Tourism Breakfast. The 2026 season showed a high return visitor rate, with 9.28 out of 10 visitors likely to return. Panelists including Erica Ashton, Alessandra Hagerty, and Elizabeth Sutherland advised businesses to use specific marketing that works well with AI search tools. Other speakers included Dave Everett of White Fang AI and David Dragas of Vidd AI.
Google Warns Agent AI Can Leak Data Without Hacking
Google Research published a report warning that personal AI agents can leak sensitive data without external hacking. Services like Meta's Muse and OpenAI's Dots store user information that can persist even after deletion, called ghost information. The report was co-authored by experts from Cornell University and UC Berkeley. Researchers recommend limiting AI permissions and requiring user confirmation before sharing sensitive data.
AI ROI Remains a Blind Spot for Financial Firms
A Sibos poll found that 61% of companies have little visibility into AI token usage despite tracking overall AI spending. Only 13% of respondents actively manage costs at the token level. Financial firms struggle to calculate AI return on investment because of unclear resource tracking. HSBC and Sumitomo Mitsui Banking Corp. shared examples of AI deployments that improved efficiency and reduced costs.
HR must demand stronger proof before AI-related layoffs
A Financial Times examination found that companies often mix AI reasons with cost-cutting and restructuring when justifying layoffs. HR should treat claims that AI causes job losses as testable ideas, not settled facts. Deloitte’s 2026 survey found only 6% of leaders said their organizations were improving human-AI work design, even though 85% called workforce adaptability critical. Gallup’s August 2026 research found 99% of CHROs saw AI as important to strategy, but 50% lacked confidence in managers guiding employees’ AI use. HR should test redesigned work, involve managers, review redeployment options and track results after layoffs to confirm whether the AI case was true.
Sources
- Utah launches health care AI push with new clinical pilots, outside evaluators
- AI may soon be in more healthcare processes in Utah under new state agreements
- OpenAI must explain action taken to stop AI hacking Australians’ private data, chair of federal inquiry says
- Optogenetics Won a Nobel, Not a Business Model
- America @ 250: Philosopher Kevin Scharp says that in the age of AI there are no questions, only requests
- Latest News In Cloud AI - Applied Digital Expands Into Finland With AI Campus Development
- There must be stricter guidelines on artificial intelligence
- Tourism Breakfast Tackles AI Strategies
- Agent AI Security — Google Warns of Ghost Data Leaks Without Hacking
- AI ROI Remains an Industry Blindspot
- AI layoffs need a stronger HR evidence standard
Comments
Please log in to post a comment.