SoftBank and OpenAI Take AI Security to Japanese Companies

Masayoshi Son has spent years telling investors that AI will change everything. On Tuesday his company began selling that argument to Japanese corporations, one vulnerability at a time. SoftBank Group, together with SoftBank Corp. and SB OAI Japan, its OpenAI joint venture in the country, launched “Patching as a Service” on June 16, an AI-powered cybersecurity offering that evaluates software for security flaws and fixes them, aimed first at roughly 3,000 large Japanese enterprises.

The service is the first concrete product to emerge from the SoftBank-OpenAI Japan venture, the partnership the two companies created to sell OpenAI technology to Japanese businesses. The joint venture structure gives SoftBank a share of the upside from OpenAI’s growth in Japan, and the patching service is its first test in the market.

Son’s pitch is urgent. Japan faces a rising tide of AI-driven cyberattacks, he said, and the country’s defenses are undermanned. Japan has long suffered from a shortage of security engineers, and the government has been pushing companies to harden their systems. The new service is designed to close the gap with machines that work at machine speed.

The product’s logic is straightforward. OpenAI’s models scan corporate code and systems, identify vulnerabilities, propose fixes and, in many cases, apply them. In practice, the service works like a continuous security audit: the AI reviews code and configurations, flags weaknesses ranked by severity, and patches what it can, with human engineers reviewing the changes before they reach production, according to people familiar with the offering.

For enterprises, the appeal is speed and coverage. AI tools can review codebases far faster than human teams, and they do not quit. The “patching as a service” model turns security from an annual audit into a continuous subscription, and it gives companies a way to respond to threats the day they are disclosed rather than the month after.

For SoftBank and OpenAI, the appeal is recurring revenue and a foothold in one of the world’s biggest corporate markets. SoftBank Corp. has sales relationships with tens of thousands of Japanese businesses, and OpenAI gains an enterprise distribution channel in a country where its direct reach is limited. The launch also marks a step for OpenAI’s enterprise strategy outside the United States.

Analysts offered a measured read. The service faces the standard problem of AI security tools: false positives and trust. A company that lets software patch its own code is making a statement about how much it trusts the model, one analyst said, and Japanese enterprises, famously cautious about new technology, will need convincing. SoftBank has not disclosed pricing, and analysts said adoption will hinge on whether the service demonstrably cuts the cost and time of security work.

The launch comes as Japan steps up its cyber defenses. The government has passed legislation for active cyber defense, ransomware attacks on Japanese companies have been rising, and executives have listed cybersecurity among their top concerns in annual surveys. The timing puts the new service in front of a market that is both worried and under-resourced.

The broader market for AI-powered security is growing quickly as attackers automate and defenders respond with automation of their own. Japanese companies have been slower than American peers to adopt AI tools, analysts said, which makes the country both a lagging indicator and a large opportunity for the venture. For now, the service begins with the 3,000 companies SoftBank already knows best, and the numbers will tell the rest of the story.

The deeper story is SoftBank’s transformation. Son has pivoted his empire from telecom toward AI investing, with major stakes in OpenAI and Arm and a portfolio of companies across the stack. The Japan venture is the bridge between SoftBank’s investing arm and its operating businesses: it turns AI investments into products sold to Japanese customers, and the patching service is the first product to cross that bridge.

For OpenAI, the venture is part of a global expansion that has included offices, data center commitments and government partnerships across Asia. The patching service is the kind of product that builds trust one enterprise at a time, and Japan is the test market. If it works, the same model can travel to other countries where SoftBank does business.

For Son, the launch is a small step toward a large claim. He has said AI will reshape every industry, and Japan’s corporate sector is the first arena where his company is selling that future directly. The service’s success will be measured in renewals, not announcements. “Patching as a Service” is a plain name for a plain strategy: sell security as a utility, meter it monthly, and make OpenAI’s models the thing Japanese enterprises cannot run without.

Related Posts

  • September 6, 2026
  • 10 views
Anthropic Moves Its IPO Filing to Late September

The bankers and lawyers running Anthropic’s initial public offering had told investors to expect the company’s registration documents as soon as this week. The calendar has moved. Anthropic now plans…

  • September 6, 2026
  • 12 views
OpenAI Quietly Revises GPT-6 Astra Scores After Launch

When OpenAI released GPT-6 Astra on Sept. 3, the launch post carried the usual furniture of a modern model debut: coding results, speed comparisons and a figure for how often…