OpenAI has acquired Torch, a small healthcare startup focused on unifying fragmented medical records, in a deal that sources say was valued at approximately $100 million in equity, according to The Information. The companies did not disclose financial terms.

As part of the acquisition, Torch’s four-person team will join OpenAI, and its technology will be integrated into ChatGPT Health, OpenAI’s newly announced service designed to help users analyze and manage their health information using AI.

The move reflects OpenAI’s growing focus on healthcare infrastructure, particularly the challenge of making scattered medical data usable by both patients and AI systems.

Torch was developing a consumer-facing application designed to aggregate a person’s medical data from multiple sources into a single, coherent system. These sources included:

The Torch team described its core technology as “a medical memory for AI”—a context engine that unifies fragmented health records so important details are not lost across systems.

In a statement announcing the acquisition, Torch emphasized the problem it set out to solve: patients often see only a fraction of their own records, while clinicians lack the time to synthesize the growing volume of data generated by modern healthcare systems and consumer devices.

The Torch founding team—Ilya Abyzov, Eugene Huang, James Hamlin, and Ryan Oman—previously worked together at Forward Health, a company known for its AI-powered doctors’ offices.

Forward Health abruptly shut down in late 2024 after raising more than $400 million, leaving many former employees and patients uncertain about the future of its technology.

Torch represented a different path for that talent and vision. With the acquisition, the team’s work will now continue inside OpenAI, rather than ending with another healthcare startup closure.

OpenAI says Torch’s technology will become part of ChatGPT Health, bringing together lab results, medications, and visit recordings into a unified experience for users.

In its announcement, OpenAI said combining Torch’s data-unification capabilities with ChatGPT opens “a new way to understand and manage your health.”

By absorbing Torch as an acqui-hire, OpenAI gains both the technical foundation and the specialized healthcare experience needed to tackle one of medicine’s most persistent problems: fragmented data spread across incompatible systems.

Q: What is OpenAI announcing?
A: OpenAI has acquired Torch, a small healthcare startup focused on unifying medical records, and will integrate its technology into ChatGPT Health.

Q: How much did OpenAI pay for Torch?
A: The companies did not disclose terms, but an unnamed source told The Information that OpenAI paid approximately $100 million in equity.

Q: What did Torch build?
A: Torch built a system it described as a “medical memory for AI,” designed to unify medical records from hospitals, labs, wearables, and health apps into a single context engine.

Q: Who is joining OpenAI?
A: Torch’s four-person team—Ilya Abyzov, Eugene Huang, James Hamlin, and Ryan Oman—will join OpenAI.

Q: How does this relate to ChatGPT Health?
A: Torch’s technology will be used to help ChatGPT Health bring together medical data such as lab results, medications, and visit recordings so users can better understand and manage their health.

The Torch acquisition underscores a quiet but critical reality about AI in healthcare: model intelligence is not the primary bottleneck—data fragmentation is.

Patients today generate more health data than ever before, but that information is scattered across hospitals, labs, insurers, wearables, and consumer apps. Without a unified view, even the most advanced AI systems struggle to provide reliable, context-aware insights.

In healthcare, progress may depend less on building smarter AI models and more on creating systems that can retain, organize, and contextualize medical history over time.

If ChatGPT Health succeeds, it could shift how individuals interact with their own health information, moving from episodic snapshots to continuous understanding. But the stakes are high: healthcare data is deeply personal, regulated, and complex. Any meaningful progress will require not just technical integration, but careful attention to privacy, consent, and human oversight.

Still, this deal suggests OpenAI is thinking beyond chat interfaces and toward foundational health infrastructure—where AI’s value comes not from replacing clinicians, but from helping people and providers see the full picture that already exists.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.