JPMorgan Chase Partners With OpenAI to Deploy Codex and Cybersecurity Models

The agreement will bring OpenAI’s coding agent and specialized cyber capabilities into the banking giant’s operations, expanding its use of AI across software development and financial security.

JPMorgan Chase has signed a strategic partnership with OpenAI to use its coding and cybersecurity models, marking another major step in the adoption of advanced artificial intelligence across banking.

OpenAI President Greg Brockman disclosed the agreement during The Wall Street Journal’s Tech conference on November 4, 2026. He said JPMorgan would use Codex, OpenAI’s coding agent, alongside its cybersecurity models to strengthen internal technology and security operations.

The partnership brings two important areas of AI adoption together: helping engineers build software and helping security teams defend critical systems.

Neither company has disclosed the financial terms, contract value, or duration of the agreement.

The announcement is significant because banks manage sensitive financial information and operate complex technology infrastructure. As a result, deploying advanced AI in these environments requires careful attention to security, governance, and operational reliability.

JPMorgan to Use OpenAI Codex for Software Development

One part of the agreement focuses on Codex, OpenAI’s AI-powered coding agent.

Codex can assist software engineers with coding tasks, including writing, reviewing, and modifying code. It can also support broader software development workflows.

For a bank the size of JPMorgan, these capabilities could help engineering teams work across large and complex technology environments.

Financial institutions maintain extensive systems covering payments, account services, trading, risk management, and regulatory reporting. Improving software development processes could help them modernize applications and support internal technology projects.

However, the announcement does not specify which JPMorgan teams will use Codex or how widely the bank will deploy it.

Brockman described the agreement as part of a wider effort to bring advanced AI into enterprise operations. He said the partnership would help engineers build, security teams defend, and employees and customers benefit from more capable AI.

This suggests the relationship could extend beyond coding alone. However, the precise applications and rollout plans have not been disclosed.

Cybersecurity Is a Central Part of the Agreement

The second major component involves OpenAI’s specialized cybersecurity models.

Cybersecurity is particularly important for large banks. Their systems handle payments, customer records, financial transactions, and connections with external institutions.

A successful attack can disrupt services, expose sensitive information, and undermine customer confidence.

AI could help security teams identify vulnerabilities, investigate threats, and improve defensive workflows. It may also help engineers understand weaknesses in software and develop fixes more efficiently.

OpenAI provides cybersecurity capabilities through its Daybreak program. The program supports approved individuals and organizations working on authorized cybersecurity tasks.

Its capabilities cover areas such as vulnerability discovery, validation, remediation, threat modelling, and security investigations.

OpenAI applies access controls and reviews applicants before granting access to advanced cybersecurity capabilities. This reflects the dual-use nature of powerful cyber models: the same technical capabilities that help defenders can also create risks if misused.

Brockman did not identify the specific cybersecurity models JPMorgan will receive. Nor did he disclose the exact security workflows covered by the agreement.

Therefore, the partnership should not be interpreted as confirmation that JPMorgan will receive unrestricted access to every OpenAI cyber model.

A Multi-Model AI Strategy at JPMorgan

The OpenAI agreement also highlights JPMorgan’s broader approach to enterprise AI.

The bank has already used OpenAI models through its internal LLM Suite platform, which provides employees with access to AI capabilities.

JPMorgan has also worked with Anthropic on cybersecurity.

In April 2026, the bank joined Project Glasswing, Anthropic’s initiative to use advanced AI for defensive security across critical software infrastructure.

The project brought together major technology and security organizations. Its objective is to identify and address software vulnerabilities before attackers can exploit them.

JPMorgan’s Chief Information Security Officer, Pat Opet, described the initiative as an opportunity to evaluate next-generation AI tools for defensive cybersecurity.

The new OpenAI partnership represents a separate relationship. It does not establish that JPMorgan has ended its work with Anthropic or selected one provider exclusively.

Instead, the developments point to a multi-model approach. Large enterprises can work with different AI providers for different technical requirements.

For banks, this strategy may offer greater flexibility. Different models can support coding, research, security analysis, and internal productivity.

However, managing multiple providers also creates challenges. Institutions must assess data handling, access permissions, model performance, integration risks, and compliance requirements across each relationship.

Why AI Cybersecurity Matters for Financial Institutions

The partnership arrives as AI changes both sides of the cybersecurity landscape.

Defenders can use advanced models to review code, investigate vulnerabilities, and accelerate security operations. At the same time, attackers can use AI to automate parts of their activities and develop more convincing attacks.

This creates pressure on financial institutions to improve their security capabilities.

Banks must protect systems that operate continuously and connect with customers, merchants, payment networks, and other financial institutions. A vulnerability in one component can have consequences across multiple services.

AI-assisted security could help teams examine complex software environments more efficiently. It may also help reduce the time between identifying a weakness and developing a response.

However, AI-generated findings still require validation. Models can make mistakes, miss vulnerabilities, or produce recommendations that need further testing.

For regulated institutions, human oversight and established security controls remain essential.

The OpenAI partnership therefore matters not simply because JPMorgan is adopting another AI tool. It reflects the growing role of advanced AI in protecting the infrastructure that supports financial services.

OpenAI Strengthens Its Enterprise Banking Push

The JPMorgan agreement also supports OpenAI’s wider effort to expand its enterprise business.

During the same conference appearance, Brockman said enterprise customers now account for the majority of OpenAI’s revenue. He contrasted this with an earlier period when enterprise represented around 30% of revenue.

He also described the JPMorgan partnership as an example of the large organizations OpenAI is increasingly working with.

Winning major financial institutions is important for AI providers. Banks require reliable technology, strong security practices, and the ability to support complex enterprise environments.

A partnership with JPMorgan could also strengthen OpenAI’s position in a market where competition is increasing.

Anthropic, Microsoft, Google, and other technology companies are also developing AI products for enterprise customers. Coding agents and cybersecurity models have become particularly important areas of competition.

For OpenAI, the deal provides another example of its attempt to move beyond general-purpose AI assistants and into specialized business workflows.

What Comes Next for JPMorgan and OpenAI?

The partnership signals a deeper role for AI in banking technology and cybersecurity.

JPMorgan will use OpenAI’s Codex and cybersecurity models, but several details remain unclear. Neither company has announced the contract value, duration, deployment timetable, or specific models involved.

The practical impact will depend on how the bank integrates these capabilities into its existing systems.

For the wider financial industry, the agreement offers a clear indication of where enterprise AI adoption is heading. Banks are increasingly exploring AI not only to improve employee productivity but also to strengthen the security of their technology infrastructure.

The next important development will be how effectively these tools support real-world banking operations while meeting strict security and governance requirements.

JPMorgan Chase signs OpenAI deal for Codex and cybersecurity models

AI cybersecurityAI in bankingbanking AICodexCybersecurity ModelsDaybreakDigital Bankingenterprise AIfinancial technologyGreg BrockmanJPMorgan ChaseJPMorgan OpenAI partnershipOpenAI