OpenAI Develops AI Operating Systems Alongside Amazon and Meta

Artificial intelligence continues to reshape various sectors, from healthcare to finance and retail, driving both innovation and new challenges. Utah recently launched a pilot program allowing residents to renew certain common prescriptions via an AI chatbot for a $4 fee, marking a first-of-its-kind initiative in the U.S. This program, a partnership between the Utah Department of Commerce's Office of Artificial Intelligence Policy and Doctronic, aims to boost efficiency and accessibility, especially in rural areas. However, physician groups like the American Medical Association warn of potential patient risks, despite Doctronic claiming 99.2% accuracy and having malpractice insurance. As AI adoption accelerates, so do security concerns. A new report highlights that organizations now manage 100 times more machine identities than human ones, straining traditional security methods. The rapid increase in "machine identities" like API keys, often lacking clear ownership, creates significant blind spots. Experts also point to "shadow AI," where employees create AI tools without IT oversight, evolving into "shadow operations" and increasing risks. Even AI-assisted coding tools, while boosting productivity, frequently generate code with security vulnerabilities, leading to data breaches as AI models prioritize speed over robust security. Major companies and industries are actively integrating AI. Hedge funds are retooling for 2026, with Bridgewater Associates developing a $5 billion AI-driven fund and expanding employee ownership. Walmart is strengthening its AI focus by adding tech veteran Shishir Mehrotra to its board, following its GenAI chatbot Sparky launch and partnership with OpenAI. Qualcomm CEO Cristiano Amon predicts "physical AI," enabling machines to sense and act in the real world, will transform industries, building on Qualcomm's success in automotive and expanding into robotics with its new AI Hub. The focus in AI computing is shifting, as observed at CES 2026, from training large models to "inference," where AI is applied in real-world scenarios. Lenovo, AMD, and Dell are introducing new inference servers to meet this demand for hybrid and edge deployments. Meanwhile, companies like Amazon, Meta, and OpenAI are developing operating systems for new AI-powered devices. These devices could allow AI agents to perform tasks like booking travel without needing specific apps, potentially disrupting traditional app business models. SAP is also launching new AI tools for retail, aiming for a complete AI-powered operating system to enhance demand planning and order reliability. Despite the rapid advancements, an editorial notes that many AI tech companies falsely claim their platforms are "first of their kind." The author, Joanna Gerber, emphasizes that agentic platforms using natural language prompts are common. Instead of focusing on uniqueness claims, companies should prioritize a strong human team that can clearly explain their product without jargon, as effective communication is crucial for standing out in the crowded AI market.

Key Takeaways

• Utah launched a pilot program for AI-powered prescription renewals for common drugs, charging a $4 fee, a first in the U.S.
• Physician groups like the American Medical Association warn of patient risks with Utah's AI prescription program, despite Doctronic's 99.2% accuracy claim.
• Organizations now manage 100 times more machine identities than human ones, creating significant cybersecurity challenges.
• AI-assisted coding tools, or "vibe coding," often produce code with security vulnerabilities, leading to data breaches.
• Hedge fund Bridgewater Associates is developing a $5 billion AI-driven fund, signaling major investment in AI for finance.
• Walmart is enhancing its AI strategy by adding Shishir Mehrotra to its board and has partnered with OpenAI.
• Qualcomm CEO Cristiano Amon foresees "physical AI" transforming industries, with Qualcomm expanding its AI chips from automotive to robotics.
• The AI computing industry is shifting focus from model training to "inference," with companies like Lenovo, AMD, and Dell introducing new inference servers.
• Amazon, Meta, and OpenAI are developing operating systems for new AI devices, potentially disrupting traditional app-based business models.
• SAP is introducing new AI tools for retail, including Retail Intelligence in SAP Datasphere and an Order Reliability Agent, to create an AI-powered retail operating system.

Utah launches AI system for prescription refills

Utah started a new pilot program that lets residents renew some common prescription drugs using an artificial intelligence chatbot. This first-of-its-kind trial allows people to get refills quickly for a $4 fee without needing to talk to a doctor. State regulators hope this experiment will help guide the future use of AI in healthcare.

Doctors warn about Utah AI prescription program risks

Utah launched a pilot program where an AI system can prescribe repeat medications without direct doctor oversight, a first in the US. While state officials say it cuts costs and helps rural areas, physician groups like the American Medical Association warn of patient risks. Doctronic, the health-tech startup behind the AI, claims 99.2% accuracy and has malpractice insurance. The program has safety checks, including doctor reviews for initial renewals and random sampling.

Utah uses AI to boost prescription renewal efficiency

The Utah Department of Commerce's Office of Artificial Intelligence Policy (OAIP) teamed up with Doctronic, an AI health platform, to make prescription renewals more efficient. This partnership allows Doctronic's AI to legally handle routine refills within a special regulatory sandbox. OAIP will evaluate the pilot for safety, patient experience, and effectiveness, tracking metrics like refill times and patient satisfaction. This is the first state-approved program letting an AI system participate in medical decision-making for renewals.

AI and machine identities strain security efforts

A new report shows that identity security faces big challenges from many non-human identities and fast AI adoption. Organizations now manage 100 times more machine identities than human ones, making security harder. While leaders expect a lot from AI, security teams see fewer positive results right now. Companies are trying to combine security platforms to better manage these identities and use AI safely.

AI security risks grow faster than human control

As AI systems become more independent, security risks are growing faster than human oversight can manage. Traditional cybersecurity methods, designed for human-paced threats, are failing against AI's speed and autonomy. The number of "machine identities" like API keys, which AI agents use to access data, is rapidly increasing and often lacks clear ownership. Experts warn that "shadow AI," where employees create their own AI tools without IT knowledge, is turning into "shadow operations," leading to more security blind spots and risks.

Hedge funds embrace AI and new business models

Hedge funds are changing their operations for 2026, driven by strong performance and the growing role of AI. Bridgewater Associates, a major firm, named Bob Prince as board chair after a record 2025, and is expanding employee ownership to retain top talent. The company is also developing a $5 billion AI-driven fund. The industry is moving towards unified AI trading platforms, like the one launched by Tradomatix, to handle complex tasks faster and more safely. Investors are now looking at how funds manage leadership, integrate AI securely, and keep skilled employees.

Securing AI coding tools prevents new cyber risks

AI-assisted development, also known as "vibe coding," offers great speed and productivity for creating software. However, these AI tools often create code that lacks important security features, leading to new vulnerabilities and real-world data breaches. This problem is worse because many people using these tools are "citizen developers" who do not have strong security knowledge. Incidents have occurred where AI-generated code led to breaches, data loss, or allowed hackers to bypass security. The main risks are that AI models focus on making code work quickly rather than making it secure, and they often miss important security details.

Physical AI will transform industries says Qualcomm CEO

Qualcomm CEO Cristiano Amon believes "physical AI," which allows machines to sense, think, and act in the real world, will be huge. This technology uses real-time sensor data to train systems, enabling robots and self-driving cars to perform complex tasks and make quick decisions. Qualcomm has become a key player in automotive technology, creating power-efficient chips for assisted and autonomous driving. Amon expects similar success in robotics, with the company's new Qualcomm AI Hub helping developers build AI applications for devices. Experts predict AI-enabled robots will soon move from specialized uses to widespread adoption.

Walmart strengthens AI efforts with new board member

Walmart is boosting its focus on artificial intelligence by adding tech veteran Shishir Mehrotra to its board of directors. Mehrotra, CEO of Superhuman and co-founder of Coda, will join the Compensation and Management Development Committee and the Technology and eCommerce Committee. Walmart leaders, including Chairman Greg Penner, highlighted Mehrotra's experience in building successful tech platforms. Mehrotra believes this era marks the most significant technological shift he has seen. Walmart has already launched its GenAI chatbot Sparky and partnered with OpenAI as part of its AI strategy.

AI computing shifts focus to inference at CES 2026

At CES 2026, a major shift in AI computing was observed, moving from training large AI models to focusing on "inference." Lenovo CEO Yuanqing Yang predicts that future AI spending will mostly go towards inference, reversing previous trends. Lenovo launched three new inference servers to lead this change, designed for various uses from data centers to industrial settings. This shift is driven by companies moving from testing AI to actually using it, and a need for hybrid and edge deployments. Businesses want to use data where it is created, control costs, and maintain data privacy. Other companies like AMD and Dell also introduced new inference servers.

SAP launches new AI tools for retail businesses

German software company SAP introduced several new AI features designed to help retail customers. These enhancements, announced before the NRF 2026 event, aim to create a complete AI-powered retail operating system. One new tool, Retail Intelligence in SAP Datasphere, will help businesses plan demand and inventory more accurately using real-time data. Another, the Order Reliability Agent, will identify and fix potential order problems and help staff answer customer questions. SAP emphasizes that AI is now essential for retailers as the industry moves towards more independent, "agentic" commerce.

AI devices challenge traditional app business models

Major tech companies like Amazon, Meta, and OpenAI are creating operating systems for new AI-powered devices, which could change how we use apps. These devices will let AI agents perform tasks for users, like booking travel or ordering food, without needing to open specific apps. This shift could impact companies like Uber and DoorDash, which rely on users staying within their apps for advertising and loyalty. While some smaller startups face resistance from companies that want direct user interaction, larger platforms are seeing partnerships with businesses like Expedia and Instacart. The future business model for AI apps is still developing, and it is unclear how companies will earn money if users interact less with traditional apps.

Most AI platforms are not truly unique

An editorial piece highlights that many AI tech companies falsely claim their platforms are "first of their kind." The author, Joanna Gerber, notes that agentic platforms using natural language prompts are common across various uses, from analytics to creative work. While technology evolves quickly, making complex tasks simple, many founders mistakenly believe their solutions are unique. The article suggests that instead of claiming to be first, AI companies should focus on having a strong human team that can clearly explain their product without using jargon. Effective communication is key to standing out in the crowded AI market.

Sources

• Utah launches first-in-the-nation trial that lets AI renew your prescription
• As Utah lets AI handle some routine prescription renewals, physicians warn of patient risks
• Utah Looks to AI to Make Prescription Renewals More Efficient
• Machine Identities, AI Adoption Challenge Identity Security Efforts
• AI Security Risks Are Outpacing Human Oversight
• Hedge Funds’ 2026 Reset: Governance Shifts, Employee Ownership, and AI-First Trading Infrastructure
• Securing Vibe Coding Tools: Scaling Productivity Without Scaling Risk
• From factory floors to offices: Physical AI is 'going to be massive'
• Walmart Beefs Up AI Push With New Tech Veteran Board Hire
• CES 2026: AI compute sees a shift from training to inference
• SAP Debuts AI Enhancements for Retail Customers
• AI Devices Are Coming. Will Your Favorite Apps Be Along for the Ride?
• Your ‘First Of Its Kind’ AI Platform Probably Isn’t The First – And You Don’t Have To Pretend It Is