Google Cloud has launched Gemini for healthcare, an AI agent for hospital and insurer administration, in US preview. It can read patient records and prepare work across systems, but clinical staff must make any changes to those records.
Google Cloud’s Gemini for healthcare connects health records, billing platforms, clinical guidelines and US Medicare datasets. Google says the AI agent can plan and carry out tasks with several steps, within guardrails and under human oversight.
The product builds on Google’s general Gemini agent and follows versions for government services, financial services and legal work. Its five ready-made skills include summarising medical records, preparing for appointments and checking insurance benefits; customers can also build their own.
For healthcare teams, a key limit is that the agent reads Epic and Oracle Health records but cannot write changes back. Answers cite their sources so staff can check them. Customers can choose models from Google, Anthropic and OpenAI, with per-seat or usage-based payment.
The preview does not yet establish whether the agent reduces administrative costs in everyday hospital use.
Read-only access keeps record changes with clinical staff, limiting how much work can be fully automated.
Support for several model providers gives healthcare buyers a choice beyond Google’s own AI models.
Written automatically by the Made in 10 Newsroom from the source above.
Google DeepMind has unveiled Gemini Robotics 2.5, an AI model for controlling robots. Google says robots using it reached a 96 percent success rate when sorting automotive parts at Mercedes-Benz factories.
Google has opened Gemini 4 Carbon to paying business and developer customers, with access also rolling out for Google AI Ultra subscribers. The AI model targets harder coding and research tasks, but longer reasoning can make each task cost more than with Argon.
Kora Systems has unveiled a Linux-based desktop operating system built around AI models and agents, with launch planned for 2027. It is designed to let agents work across applications and files while retaining context between tasks.