OpenAI Acquires Torch: Building the ‘Unified Medical Memory’ for ChatGPT Health”

OpenAI has acquired healthcare technology startup Torch in a deal reportedly valued between $60 million and $100 million, significantly expanding the capabilities of ChatGPT Health. The acquisition is aimed at building a “unified medical memory” for AI—a system that can consolidate fragmented patient data from hospitals, labs, wearables, and consumer health services into a single, context-aware framework.

Torch was designed to address one of healthcare’s most persistent challenges: data silos. Its technology acts as a context engine that aggregates health information from multiple sources, including clinical systems and consumer platforms like genetic testing services and wearables, allowing AI to interpret a person’s health history as a continuous whole rather than isolated data points. OpenAI plans to integrate this capability into ChatGPT to better support the roughly 40 million users who already ask health-related questions on the platform each day.

As part of the acquisition, Torch co-founders Ilya Abyzov and Eugene Huang will join OpenAI to lead integration efforts. The move comes amid an “agentic shift” among major AI players, with OpenAI, Anthropic, and Google all racing to build AI systems that can reason over long-term, personalized context rather than provide one-off responses. By embedding Torch’s technology, OpenAI aims to ensure that critical medical information isn’t lost and that health-focused AI agents can operate with deeper continuity, awareness, and usefulness.