OpenAI grants Japan's megabanks access to GPT-5.5 to defend against AI-powered cyber threats

OpenAI gives Japan banks access to latest model, Japan's finance minister says

TOKYO, May 29 (Reuters) - Some Japanese financial institutions were granted access to U.S. AI giant OpenAI's GPT-5.5 model to defend against cyberattacks, Japanese Finance Minister Satsuki Katayama said ⁠on Friday. "This is a welcome development and a big step forward in strengthening Japanese financial institutions' ability to defend against cyberattacks," Katayama told reporters after a meeting in Tokyo with Jason Kwon, OpenAI's chief strategy officer. Katayama did ⁠not disclose the name of the financial institutions. The Nikkei newspaper reported on Thursday that Japan's three biggest banks - MUFG Bank, ⁠Sumitomo Mitsui Banking Corp and Mizuho Bank - were expected to gain access to OpenAI's ⁠latest model, available only to trusted partners and believed to ⁠be on a par with that of rival Anthropic's Claude Mythos. Reporting by Leika Kihara; Editing by Himani Sarkar Our Standards: The Thomson Reuters Trust Principles., opens new tab

OpenAI Gives Japan's Biggest Banks A Cybersecurity Weapon - Microsoft (NASDAQ:MSFT)

OpenAI Targets Japan's Financial Core Japan's three largest banks are at the center of this deployment. MUFG Bank, Sumitomo Mitsui Banking, and Mizuho Bank are expected to receive access, Nikkei reported Thursday. OpenAI is delivering the model through its "Trusted Access for Cyber" program, a framework designed to ensure only verified defenders can use the tool. The urgency behind this deal is real. Anthropic's Claude Mythos model, released in April 2026, demonstrated the ability to autonomously find and exploit zero-day vulnerabilities. Zero-day flaws are security gaps that developers have not yet identified or patched. Since Mythos went public, financial institutions worldwide have scrambled to build defensive countermeasures. Japan's megabanks are not waiting. Moreover, the deployment sits inside a broader US-Japan cybersecurity agreement covering 15 critical infrastructure sectors. A public-private working group of 36 member entities now coordinates that effort. Notably, both OpenAI and Anthropic sit on that group alongside Japan's megabanks and other key stakeholders. OpenAI Is Building an Asian Footprint, Not Just Closing Deals Japan is not a standalone win. OpenAI is constructing a regional presence across Asia that goes well beyond single-country agreements. That matters because sovereign AI partnerships create long-duration infrastructure demand. Cloud compute, cybersecurity spending, and regulatory integration become recurring revenue layers rather than one-time software sales. Together, the Japan and Singapore moves reveal a pattern. OpenAI is not simply selling access to its models. It is embedding itself into the economic infrastructure of allied governments across Asia. The company's broader "OpenAI for Countries" initiative targets at least 10 national or regional partnerships in its first phase, and Asia is becoming the most visible proving ground. While OpenAI Expands Abroad, Anthropic Builds Its U.S. Enterprise Base Anthropic is taking a different path. The San Francisco-based company reported a run-rate revenue of $30 billion as of April 2026, driven largely by U.S. private enterprise customers. CEO Dario Amodei noted that roughly 80% of Anthropic's revenue comes from enterprise clients. That base is deep and difficult to displace. The company has faced headwinds in Washington. President Trump directed federal agencies to cease using Claude in February 2026, after Anthropic refused to allow its models to be used for autonomous weapons systems or mass domestic surveillance. Anthropic subsequently filed a lawsuit arguing the government retaliated against protected speech. Nevertheless, its private enterprise business has continued to grow, and Anthropic remains part of the Japan cybersecurity working group, signaling that its international footprint is expanding, just more quietly. What This Means for Investors Benzinga Disclaimer: This article is from an unpaid external contributor. It does not represent Benzinga's reporting and has not been edited for content or accuracy. Market News and Data brought to you by Benzinga APIs To add Benzinga News as your preferred source on Google, click here.

OpenAI to Grant Some Japanese Banks Access to New Models

TOKYO--OpenAI has agreed to give some Japanese banks access to its new artificial-intelligence model, Japan's finance minister said, a move that could bolster the nation's cyber-defense. "With frontier AI now recognized as a threat, this step could serve as a welcome catalyst to strengthen cybersecurity within Japan's financial sector," Finance Minister Satsuki Katayama told reporters Friday after meeting with OpenAI executives. Being able to use the model, known as GPT-5.5 Cyber, could allow Japan to better scrutinize systemic risks, pinpoint gaps in information security and implement countermeasures. Japan's financial sector has been closely tracking the emergence of frontier AI models that have become a focal point of global concern due to their ability to autonomously identify software vulnerabilities. Models like Anthropic's Claude Mythos and OpenAI's GPT-5.5 Cyber can discover and exploit flaws at a velocity that far outpaces human capabilities, raising alarm that unleashing them poses an existential threat to financial systems with outdated cyber defenses. Some Japanese financial firms will also gain access to Mythos after an approval from the Trump administration, Katayama said.