Hello AI Enthusiast,
This week AI moved further into the places where it can do the most good. ChatGPT connected to millions of US medical records to help doctors, the big labs put fresh focus on security as their models grow more capable, and Europe brought ChatGPT under the same rules as the largest platforms. Let's get into it.
The Big Picture 🔊
Doctors can now pull Epic records into ChatGPT
OpenAI has connected ChatGPT for Healthcare to Epic, the platform that holds records for more than 325 million US patients. Access is limited to authorized clinicians at organizations that turn it on, with role-based access and audit logs, and it is read-only, so it cannot change the chart. It is built for hospitals and clinical teams, not personal accounts. A separate plugin links ChatGPT to nine official sources like PubMed and ClinicalTrials.gov.
|
||
|
This is a big deal. Epic holds the records for most of the country, so connecting ChatGPT to it puts AI right next to almost every US patient's chart. Rather than compete with ChatGPT on the chatbot experience, Epic is choosing to plug in, and recent data shows why that is smart: in Offcall's 2025 physician survey, doctors reached for purpose-built clinical tools like OpenEvidence (44.9%) far more than general ChatGPT (15.6%), with Epic's own AI near the bottom of the list. |
Healthcare teams are getting powerful AI tools, but a tool only helps if people know how to use it and where it should not be trusted yet. Closing that gap between having AI and using it well is exactly what our Corporate AI Training does, from spotting where AI adds real value to building approved, reliable workflows your team can rely on.
OpenAI and Anthropic put security front and center
OpenAI and Anthropic both published detailed accounts of recent security incidents, including an unreleased OpenAI model that broke out of its test environment, reached the internet, and exploited a live vulnerability on its own. Each confession came with a cure. OpenAI is preparing Astra, a model built for cyber defense that may reach its "critical" capability level, while Anthropic rolled out Enterprise Frontier Safeguards so large customers can watch for misuse on their own cloud.
|
||
|
Ironically, the same companies building highly capable models are also building the products to secure them. These models have outgrown the safety tests designed for them, so the labs are writing new capabilities and new safeguards at the same time. |
Europe now treats ChatGPT like a Search Engine
The European Commission has designated ChatGPT as a Very Large Online Search Engine under the Digital Services Act, the first standalone AI service to fall under the EU's strictest tier. Reddit and Roblox were named the same day. All three crossed 45 million monthly EU users, with ChatGPT reporting about 159 million. OpenAI now has four months to add annual audits, ad transparency, risk reviews, and data sharing with regulators.
|
||
|
This is the first time an AI chatbot has been placed in the same regulatory box as Google, and it will not be the last. The label follows the web-search feature rather than the brand, so as tools like Claude add browsers and grow their audience, expect more of them to join the list. |
Bits and Bobs 🗞️
Anthropic's Claude desktop app added a built-in browser that reads pages and fills forms without touching your personal browser data.
A federal judge ruled that the Pentagon's "supply-chain risk" label on Anthropic was unlawful retaliation.
Google introduced Gemini 3.5 Transcribe, a speech-to-text model with real-time, multilingual transcription.
OpenAI said it will end its contract with Cursor by November 2026, after SpaceX acquired the coding tool.
Anthropic previewed the Model Hardware Standard, letting Claude operate lab and manufacturing devices through one interface.
OpenAI reported that ChatGPT Ads reached a $1 billion annual revenue run rate in under 200 days.
Anthropic released Claude Fable 5.1 and Mythos 5.1, with added safeguards for cybersecurity and life sciences.
Salesforce and Anthropic announced Claudeforce, connecting Claude to Salesforce data and workflows.
McKinsey published its 2026 Global State of AI survey.
OpenClaw shipped its largest update, with easier setup and support for existing ChatGPT subscriptions.
That's a wrap on our newsletter!
If you're looking to build AI skills in your team, check out our Corporate Training. Every program is tailored to your team and built around hands-on practice.
Catch you next week! 👋
