MUFG rolls out ChatGPT Enterprise to roughly 35,000 employees in AI-native push with OpenAI

Mitsubishi UFJ Financial Group is pushing ChatGPT Enterprise across Mitsubishi UFJ Bank in a phased rollout that OpenAI says now reaches approximately 35,000 employees, part of a broader effort the bank calls becoming an "AI-native" company.

What's new

OpenAI's newsroom detailed the rollout in a case study published July 7, 2026. Mitsubishi UFJ Bank, the core banking arm of MUFG, began a phased deployment of ChatGPT Enterprise in 2026 so that roughly 35,000 employees could use it as part of their daily work, alongside a push to build new AI-powered customer experiences.

Tadashi Yamamoto, Group CDTO at MUFG, framed the partnership choice directly: "OpenAI shared our vision of becoming an AI-native company. With its frontier technology and models, we saw a partner that could help us turn that vision into reality."

Kohei Shimano, Managing Director and Head of the Artificial Intelligence and Solutions Department at Mitsubishi UFJ Bank, said the deciding factors for the enterprise-wide rollout were broad applicability and secure use, since many employees already had informal familiarity with ChatGPT and enterprise-grade security controls were needed to meet the governance requirements expected of a financial institution.

Context

MUFG is one of Japan's largest financial groups, spanning banking, trust banking, securities, cards, and leasing, serving individuals and businesses in Japan and internationally. The relationship with OpenAI dates to October 2024, when MUFG began working with the company to modernize financial operations and improve efficiency using generative AI. The 35,000-employee rollout marks the transition from that early collaboration into an organization-wide deployment, one of the larger disclosed enterprise ChatGPT rollouts in the banking sector.

OpenAI frames the engagement as more than a productivity tool rollout. Per the case study, MUFG's transformation combines top-down commitment from leadership with bottom-up enablement, encouraging employees to build AI into daily workflows rather than treating it as a specialist tool.

Why it matters

A 35,000-employee deployment inside a systemically important financial institution is a meaningful data point for enterprise AI adoption in a heavily regulated industry, where security, auditability, and governance requirements typically slow rollouts far more than in less-regulated sectors. Shimano's framing — that broad applicability and secure use were the deciding factors, not raw capability — signals what large financial institutions are actually screening for when picking an AI vendor at scale.

The deal also underscores OpenAI's enterprise strategy of anchoring large, multi-year relationships with major regulated institutions (alongside deals like its Australian Payments Plus partnership announced the same day) as a hedge against consumer-product competition and a way to lock in switching costs before rivals like Anthropic or Google establish similar footholds. As Yamamoto put it, "I believe AI will fundamentally change the nature of finance" — and MUFG's bet is that being an early, full-scale enterprise adopter gives it a head start on that shift rather than leaving it to catch up later.