$500 Billion Crypto Giant Launches Medical Large Model, Outperforming Google

Recently, Tether, known for issuing the world's largest stablecoin USDT, launchedA series of medical language models designed to run on smartphones, wearable devices, and edge devicesQVAC MedPsy, and open-sourced on Hugging Face.

The ModelDemonstrates strong performance in multiple medical reasoning and diagnostic benchmark tests:This lightweight model, with only 1.7 billion parameters, outperformed Google's MedGemma-27B, which is 16 times larger in scale; moreover, the 4-billion-parameter version used 3.2 times fewer tokens than mainstream models, maintaining high output quality while reducing computational costs.In addition, both the 1.7B and 4B models launched this time are in the quantized GGUF format, small enough to run on standard consumer-grade hardware without transmitting sensitive information through the cloud. This is also the biggest selling point of the model.Keep the user's health data, symptom records, medication reminders, mental state, and wearable device data as much as possible on the local device, without uploading all of it to OpenAI, Google, or a certain medical SaaS platform.

Alphabet's Isomorphic Labs Raises Over $2 Billion for Medical AI Project

Isomorphic Labs, a subsidiary of Alphabet, is reportedly in talks to raise over $2 billion in its latest funding round. The company, which leverages artificial intelligence to expedite drug discovery, is expected to be led in this round by Thrive Capital, an investor in OpenAI Group PBC and an existing backer of Isomorphic Labs. Alphabet is also said to be participating in the investment.

Isomorphic Labs spun off from Google's DeepMind machine learning research lab in 2021. The company is led by Demis Hassabis, co-founder of DeepMind. Hassabis, along with another DeepMind employee, was awarded the 2024 Nobel Prize in Chemistry for developing an artificial intelligence model called AlphaFold 2. This algorithm can automatically predict protein structures, a task that traditionally required years of manual work to accomplish. The subsequent release of AlphaFold 3 improved accuracy by over 50% in certain research tasks. It excels particularly in analyzing small molecules, which are biological building blocks and the foundation of many drugs.

The flagship product of Isomorphic Labs is an artificial intelligence system called the "Isomorphic Labs Drug Design Engine" (abbreviated as IsoDDE). This system aims to accelerate drug research and development projects by automating manual research workflows. Isomorphic Labs states that, in certain tasks, IsoDDE performs significantly better than AlphaFold 3.

Bayesian Health Receives FDA's First Continuous AI Sepsis Monitoring Clearance

On May 12, Bayesian Health announced that its sepsis early warning device had received FDA 510(k) clearance.This solution monitors hospitalized patients, detects deterioration early, and alerts for sepsis.. This approval follows the FDA Breakthrough Device Designation previously granted to Bayesian, representing a significant advancement in clinical artificial intelligence and providing healthcare professionals with verified, trustworthy, and real-time intelligent support that can be used at the bedside.

"Pre-screening can buy precious time, and time is the key to changing the prognosis of sepsis. Once clinicians suspect sepsis, the clock has already started ticking — often several hours or even days have already passed," said Dr. Suchi Saria, founder and CEO of Bayesian Health and director of the Johns Hopkins AI and Medicine Lab. "There are currently no other approved tests or devices that can monitor for sepsis before clinicians become suspicious."

Anthropic Focuses $200 Million in Funding on Healthcare Sector

Anthropic has committed $200 million in funding to the Gates Foundation and will apply its Claude chatbot to related projects, with a focus on healthcare. Resources will be invested in the research and development of new drugs and vaccines, and assisting governments around the world in better utilizing health data, with particular attention to low- and middle-income countries.

Anthropic stated on May 14: "As part of our work in the field of healthcare intelligence, we will develop connectors (enabling Claude to directly access other platforms and tools), benchmarking, and evaluation frameworks to help researchers, developers, and governments better understand the performance of AI systems in healthcare-related tasks. Together, we will explore how AI can better assist frontline healthcare workers and patients in addressing diagnosis, treatment, and medical decision-making."

OpenAI Adds Codex Support for HIPAA-Compliant Medical Applications

OpenAI is expanding Codex's capabilities by adding mobile access and supporting HIPAA-compliant usage in specific enterprise healthcare environments.

OpenAI announced that its AI programming and workflow assistant, Codex, is now available through the ChatGPT mobile app. Users can monitor and manage ongoing tasks via their phones while the system operates on connected laptops, remote machines, or within enterprise environments. Eligible ChatGPT Enterprise workspaces can now utilize Codex in HIPAA-compliant local environments via the command line, integrated development environments, and application-based local setups. The new features aim to help healthcare organizations "support patient care and operational workflows with greater speed and confidence."

OpenAI Faces Wrongful Death Lawsuit, ChatGPT Health Service Asked to Halt

A teenager died of a drug overdose after consulting ChatGPT, and the family is now suing OpenAI and requesting the court to suspend the ChatGPT Health service.

Court documents filed on May 12 in San Francisco County Superior Court in California show that Leila Turner-Scott and Angus Scott claim their 19-year-old son, Sam, died from an overdose after ChatGPT allegedly instructed him on how to mix the drug alprazolam (Xanax) with the herbal supplement kratom. The plaintiffs are seeking financial compensation and a suspension of the ChatGPT Health service until its safety is independently verified.

ChatGPT Health Allows Users to Upload Medical Records to Aid Its Responses, the Service Launched in a Limited Test Format in January. Currently, the Service Has a Waitlist. OpenAI Stated That Doctors Will Provide Ongoing Feedback on the AI’s Responses. Reportedly, Over 40 Million Americans Consult ChatGPT Daily on Health-Related Issues. However, OpenAI and Other Chatbot Developers Are Facing an Increasing Number of Wrongful Death Lawsuits Due to the Recommendations Provided by Their Artificial Intelligence.

Stanford Medical Center Uses Environmental AI Tools to Process Over 1 Million Medical Records

Troy Foster, Director of Digital Health at Stanford Medical Center, recently wrote that the center has used DAX Copilot (now renamed Dragon Copilot) to generate 1 million clinical records. "A total of 1,001,018 records, with more than 1,600 clinicians using it daily," he wrote.

Dragon Copilot is an ambient AI documentation platform launched by Microsoft, designed to automatically generate medical records during patient visits, thereby reducing the administrative burden on clinicians. Multiple healthcare institutions have adopted this technology as part of broader initiatives to optimize workflows, improve efficiency, and alleviate clinician burnout.

AMA Releases Policy Framework to Combat AI Deepfake Impersonation of Doctors

The American Medical Association (AMA) has introduced a comprehensive framework, developed by AMA's Center for Digital Health and Artificial Intelligence, aimed at protecting physicians from unauthorized AI-generated deepfake videos.

The framework is based on seven policy principles: treating the identity of physicians as a protected right; prohibiting deceptive impersonation of doctors; informed, proactive, and revocable consent; mandatory transparency and labeling; shared responsibility in preventing impersonation; enforcement and feasible remedies; and minimizing administrative burdens. The American Medical Association (AMA) uses the term "augmented intelligence" when referring to artificial intelligence, emphasizing its supportive role in the medical field.

Australia "My Health Record"Project Receives Additional $400 Million in Funding"

The Australian government will invest more than 400 million US dollars to upgrade the national digital health record system "My Health Record", while working on requiring the sharing of more critical health information and developing a new national digital child health record system.

In the 2026-2027 fiscal year budget, the Australian government allocated AUD 598.3 million for the continued operation and upgrade of the "My Health Record" system over the next two years. The Department of Health, Disability, and Aged Care stated in a media release that further improvements to the "My Health Record" system would enable patients and healthcare professionals to "access more reliable and timely health information." The Australian Digital Health Agency (the peak industry body in Australia's digital health sector) said that this year’s budget "further solidifies the trusted and secure development and application of Australia's digital health sector." The government stated that this move will make it more convenient for over 24 million Australians to use "My Health Record."

India and Japan Collaborate in Health Research, Quantum Science, and Artificial Intelligence Fields

Recently, following a high-level meeting held in Delhi, India and Japan formally confirmed new bilateral cooperation in fields such as health research, quantum science, and artificial intelligence. Japanese Minister of State for Science and Technology Policy, Komi Onoda, led a delegation to meet with Dr. Jitendra Singh, India's Federal Minister of Science and Technology, where they exchanged formal agreements covering medical devices and quantum technology.

Japan's Agency for Medical Research and Development, the Indian Council of Medical Research, and India's Ministry of Science and Technology signed a Memorandum of Understanding on cooperation in the fields of health and medical devices. Additionally, Japan's Cabinet Office and India's Ministry of Science and Technology signed a Letter of Intent on collaboration in quantum science and technology. The delegations from both sides discussed cooperation under the framework of India’s National Quantum Mission, covering areas such as quantum computing, communication, sensing, and secure networks. Japan introduced its Quantum Innovation Center and expressed interest in aligning it with Indian institutions.

Singapore Launches Latest AI Tool to Predict Risk of Diabetic Amputation

Singapore General Hospital, Singapore Health Services (SingHealth), and the Ministry of Health's Office for Healthcare Transformation have jointly developed an artificial intelligence model aimed at predicting the risk of lower limb amputation in diabetic patients. The model can provide warnings three to five years before the onset of foot ulcers and infections.

LEA-Net Utilizes Anonymous Data from Over 830,000 Patient Records of Singapore Health Group (SingHealth) — Including Demographic Information, Clinical Conditions, and Medical Examination Results — to Categorize Patients into Low-Risk and High-Risk Groups, Supporting Earlier Referrals or Interventions. A Media Release Stated That in a Validation Study Using 250,000 Patient Records, the Model Achieved Sensitivity Close to 80% and Specificity Close to 90%. The Research Team Plans to Further Validate the AI Model Through a Pilot Study Involving Patients from the Singapore Health Group (SingHealth) Diabetes Registry.

The First Artificial Intelligence Alliance for Women's Health Established

On May 12, maternal and child health organization Willow Innovations, Inc. and women's health artificial intelligence platform Ema EQ jointly announced the establishment of the "Women's Health Artificial Intelligence (WHAI) Alliance." This alliance is the first industry organization dedicated to formulating common benchmarks, ethical standards, and transparent evaluation methods for artificial intelligence in the field of women's health. Reports indicate that with the establishment of this alliance, AI tools are expected to meet accuracy standards, provide data that helps bridge the data gap in women's health, and ensure transparency in their data sources.

‌Willow InnovationsIsomorphic Labs is an American technology company focused on women's health, founded in 2014 and headquartered in Mountain View, California. The company is renowned for its innovative wearable breast pump and is committed to enhancing the experience and support for breastfeeding mothers through the integration of hardware and software.Ema EQ is an artificial intelligence system specifically designed for women's health. The founding members of the "Women's Health Artificial Intelligence Alliance" include Willow and Ema EQ, as well asWomen's Health Application ‌Clue‌ 、Health and Productivity PlatformThrive Global, wearable device company Oura.

AI Tools Have Long Been Integrated Into Women's Health Management. However, These Models Are Trained With A Severe Lack Of Female Data, Deployed Without Adequate Oversight, And Validated Against Benchmark Standards That Initially Did Not Consider Women. The Establishment Of The "Women's Health AI Coalition" Aims To Fill This Gap Before Such Conditions Become The Norm. It Is The First Industry Organization Dedicated To Developing Common Benchmarks, Ethical Standards, And Transparent Evaluation Methods For AI In Women’s Health.