OpenAI Partners With JPMorgan on Codex and Cybersecurity AI

OpenAI is expanding its work with JPMorgan Chase into coding and cybersecurity, adding its Codex agent and specialized cyber models to a bank relationship that already includes OpenAI models in an internal AI platform.

OpenAI President Greg Brockman disclosed the partnership Wednesday, Nov. 4. The agreement covers engineering and cybersecurity, along with broader work across the bank. JPMorgan plans to use Codex and the specialized cybersecurity models internally to strengthen security.

Brockman said trust and regulatory standards are central to this work. He also said enterprise now accounts for a majority of OpenAI’s revenue, compared with about 30% previously, a shift covered in earlier coverage of his comments on OpenAI’s business.

JPMorgan has used OpenAI models through its LLM Suite since 2024. The platform also offers Anthropic models and is available to hundreds of thousands of employees. By early 2026, it had about 150,000 active users each week.