OpenAI partners with JPMorgan Chase to bring AI to cybersecurity and engineering

OpenAI has a new enterprise customer, and it is not a startup with a ping-pong table. The company announced a partnership with JPMorgan Chase to apply advanced AI across engineering, cybersecurity, and broader enterprise work.

OpenAI president Greg Brockman revealed the deal at the Wall Street Journal Tech Conference on November 4, 2026.

What the partnership actually covers

The core of the arrangement involves two OpenAI products. The first is Codex, the company’s coding agent. The second is a set of specialized cybersecurity models.

JPMorgan plans to deploy both tools internally to strengthen its security. The stated goal is wider still: supporting a transformation that spans the entire enterprise.

Brockman framed the deal around trust. He stressed that meeting regulatory standards is central to this kind of work, and he positioned JPMorgan as a key institution on that front.

The partnership also reflects a shift inside OpenAI itself. Enterprise revenue has grown from approximately 30% of the company’s business to a majority share.

JPMorgan has been here before

This is not a first date. JPMorgan has used OpenAI models since 2024 through its internal LLM Suite platform.

LLM Suite gives the bank’s staff access to models from both OpenAI and Anthropic. It has reached hundreds of thousands of employees, and by early 2026 it supported about 150,000 active users each week.

The bank says the results have been tangible. Its AI initiatives generated roughly $2 billion in annual savings by early 2026, alongside substantial productivity gains.

JPMorgan has also kept its options open on the security side. Since April 2026, it has worked with Anthropic on Project Glasswing, an effort focused on defensive cybersecurity using AI.

What this means for banks, OpenAI, and the AI race

For Anthropic, the picture is more nuanced. JPMorgan already works with it on Project Glasswing and offers its models through LLM Suite. The new OpenAI arrangement does not replace that relationship, but it does mean the two labs are competing for the same bank’s cybersecurity budget.

The roughly $2 billion in annual savings is the number rivals will notice first.

Brockman’s emphasis on trust and regulatory standards suggests both companies know this. Regulators will likely want to see how the bank validates AI-generated code and how its cyber models are monitored.