OpenAI expands into healthcare with new hires and tools

https://lnkd.in/dWatdZcx From the article, "OpenAI Makes a Play for Healthcare": "When using AI, no matter how informed we may be about a topic, people tend to value the model’s recommendations over their own beliefs. This bias is made even more dangerous by the fact that AI is inherently a black box: we have no idea why or how it gets to the conclusions it does, making it harder to understand where the reasoning could have gone wrong and whether you should trust the model."

OpenAI is officially going all in on "healthcare AI" and things are really heating up! With new leadership hires and partnerships, including its GPT-5 announcement highlighting health capabilities, OpenAI is making a big bet that AI can explain lab results, decode medical jargon, and support clinicians and patients alike. Early studies like Stanford’s suggest AI can SOMETIMES outperform physicians in diagnosis, but real-world failures remind us of the risks: in healthcare, even a small error rate can be catastrophic! For clinicians and healthcare leaders, the key takeaway is balance: AI can be an incredible tool for efficiency and patient empowerment, but it must be used with clear safeguards, human oversight, and ongoing validation. The technology won’t ever replace doctors, but it will transform how patients interact with their health data and how providers make decisions. The question isn’t if AI will reshape healthcare... it’s how safely and responsibly we integrate it. read the full article here: https://lnkd.in/gMn2bh5F #HealthcareAI #HIPAA #HealthTech #PatientSafety #AIethics #DigitalHealth #OpenAI

The future of healthcare will be co-written by AI and the messaging around it will matter just as much as the models: https://lnkd.in/e-5FQ-EC Here’s what’s on my mind: - How will patients react to AI-enabled care? - Who controls the narrative around risk, bias, and equity? - How do we communicate AI's value without overhyping its capabilities? What’s your take on OpenAI’s move?

GPT-5’s launch came with a clear signal: OpenAI sees healthcare as a top use case. Consumers have long turned to Google or WebMD before (or after) visiting their doctor. AI makes those interactions more conversational—and perceived as more trusted. That creates an opportunity to engage patients in new ways, but it also raises important questions: 1. What happens when patients upload identifiable health data to a public model? 2. If AI reassures someone that “everything looks fine” and it isn’t, who’s liable? 3. How do we ensure safeguards, transparency, and patient education keep pace with adoption? AI will be part of healthcare’s future. The challenge ahead is building the systems, standards, and accountability that ensure it’s safe, equitable, and supportive of the people—clinicians and patients—it serves. https://lnkd.in/eGq9mAzM

OpenAI is moving from powering others’ tools to building its own healthcare applications. With GPT-5’s strong performance on clinical benchmarks and the release of HealthBench, the company is positioning healthcare as a core pillar of its AI strategy. This marks a shift toward end-to-end solutions for clinicians and consumers, entering a space already shaped by Microsoft, Palantir, Abridge, and others. OpenAI’s direct entry signals the next stage of competition and innovation in healthcare AI. https://archive.ph/QIYjw

While 88% of health systems are using AI, only 18% have mature governance. https://lnkd.in/esPhiut5

How to develop AI policies that work for your organization’s needs What to include in an AI policy: The AMA toolkit includes a model AI policy document that health care organizations can download and modify to align with their existing governance structure, roles, responsibilities and processes. Health systems should also be aware of state and federal laws that could impact AI’s use. At minimum, an organization’s AI policy should articulate: ✅ Definitions for relevant terms such as generative AI and machine learning. ✅ AI risks, including the risks associated with a lack of transparency, patient safety and data privacy and security. ✅ Permitted uses of approved and publicly available AI tools. This includes describing permitted-use cases, such as using AI for developing drafts of marking materials and research summaries. ✅ Prohibited AI uses, such as entering patients’ personal health information into publicly available AI tools. ✅ Permitted uses of approved AI tools, such as requirements and guidelines that all team members should follow when using AI tools. ✅ Governance, evaluation and approval processes for AI tools. ✅ Description of AI accountability and oversight, including risk assessment and regulatory compliance. ✅ Policy for how long AI generated information and a patient visit recordings will be retained. ✅ Transparency, including guidelines on when and how clinicians and patients should be made aware that AI is being used. ✅ Training, for example, incorporating AI training in the annual and ad hoc training program for everyone in the organization who uses AI. https://lnkd.in/dQJEb8nK

A recent survey of 233 health system executives found that 88% of use #AI in some form, but only 18% have a mature governance structure and full AI strategy. Why getting departments to use AI in the same way, with the same safeguards, is good for healthcare: https://lnkd.in/gSuH7ACG

We surveyed 233 healthcare executives and found that 88% of health systems are using AI, but only 18% have mature governance. Click here for the full report: https://lnkd.in/gWxJPxnk https://lnkd.in/gSuH7ACG

An Epic vision for using AI – but will it work in the real world? The company plans to infuse artificial intelligence throughout its applications, but larger questions remain for customers and the wider industry. https://lnkd.in/g5FCqhtj