OpenAI reportedly now has 10M users for Codex for coding and ChatGPT Work

Photo: Kevin Dietsch/Getty Images

OpenAI now has 10 million people using two of its agent products, Codex for coding and ChatGPT Work, doubling usage from earlier this month, Seeking Alpha said, citing a Bloomberg News report.

Large health systems are using ChatGPT for Healthcare since OpenAI launched the product in January to help providers deliver care for patients while supporting HIPAA compliance requirements.

ChatGPT for Healthcare was rolled out to several health systems, including AdventHealth, Baylor Scott & White Health, Boston Children's Hospital, Cedars-Sinai Medical Center, HCA Healthcare, Memorial Sloan Kettering Cancer Center, Stanford Medicine Children's Health and University of California, San Francisco.

In May, OpenAI expanded Codex support for HIPAA-compliant healthcare use.

OpenAI competes with Anthropic on the development of AI agents. Anthropic's answer to OpenAI's products has been the release of Cowork. Another competitor is China.

China just released two AI models that claim to rival OpenAI and Anthropic and they're free, according to Entrepreneur.

The U.S. and China will reportedly hold discussions over AI in September.

OpenAI's head of Strategic Futures Dean Ball drew the ire of top Trump administration officials over the weekend when he said the government shouldn't ban open source software but sow "fear, uncertainty and doubt (FUD)" to scare companies away from using the Chinese models, according to the Times of India.

The news comes as Anthropic and OpenAI are preparing to go public at valuations near $1 trillion and expect large profit margins, according to RealClearMarkets. Lobbyists are urging Washington to treat open-model AI as a security threat, the report said.

Yahoo reported in June that OpenAI was forced to pause the internal deployment of one of its experimental AI models after it began looking for ways to break free of its constraints. The artificial intelligence model that was built to operate autonomously for hours or days was able to "learn the blind spots" of security systems designed to contain it and "work around [them] to achieve its goals," according to the Independent.

The testing reportedly took place inside what researchers refer to as a sandbox – a tightly controlled environment meant to isolate software from the outside world.

Email the writer: [email protected]