SoftBank OpenAI launch AI-driven cybersecurity service to defend Japan critical infrastructure

SoftBank OpenAI Roll Out Patching as a Service for Japan Critical Infrastructure

SoftBank Group and OpenAI have launched a cybersecurity service designed to shield Japan's most vital companies from the escalating threat of AI-powered cyberattacks. The product, called Patching as a Service, will initially target approximately 3,000 firms operating Japan critical infrastructure including airports, power systems, and transportation networks1

1

. SoftBank founder and CEO Masayoshi Son framed the threat in urgent terms during a Tuesday presentation to enterprise clients in Tokyo, calling Japan's vulnerability to AI-enabled breaches "a crisis" and comparing modern attacks to machine guns replacing the rifle shots of the past2

2

.

The launch represents a significant deepening of ties between SoftBank OpenAI, with SoftBank having committed cumulative investments of $64.6 billion in OpenAI by the end of 20263

3

. Son emphasized the defensive mission, stating "I feel it is our duty" and repeatedly referring to criminal attackers as "the bad guys" while positioning OpenAI's technology as "the new weapon" to defend against cyber threats1

1

3

3

.

How AI-Driven Cybersecurity Works in Practice

Patching as a Service delivers AI-driven cybersecurity through vulnerability assessments, remediation planning, and implementation support, though the actual application of fixes remains in human hands4

4

. The service runs AI models to probe systems for weaknesses, then helps organizations plan how to address those gaps and advises on execution. Expert human teams still handle the prioritizing, planning, and final implementation—a critical distinction for systems where a single missed flaw could disable a power grid or airport2

2

.

SoftBank tested this approach on its own internal systems using OpenAI's cyber models and reported "promising results," gaining operational experience that will inform the product rollout4

4

. SoftBank President and CEO Junichi Miyakawa said the company would "confront the increasingly sophisticated cyber threats targeting Japan's critical infrastructure" by leveraging practical expertise acquired through this internal testing4

4

. The team working on the product rollout currently numbers around 50 people but is set to expand to approximately 1,000 people3

3

.

Joint Venture Brings AI Models to Japanese Market

The cybersecurity service will be delivered through SB OAI Japan GK, a 50:50 joint venture established in November 2025 between SoftBank's domestic telecoms arm and OpenAI to develop and exclusively market AI services for the Japanese market1

1

3

3

. The partnership combines OpenAI's AI models with SoftBank's operational capabilities and local market knowledge4

4

.

Sam Altman, OpenAI's CEO, was scheduled to attend the Tokyo launch but appeared only in a brief video, explaining that his daughter was born earlier than expected. Mark Chen, OpenAI's chief researcher, represented the company in person instead1

1

. Altman stated that "AI is transforming cybersecurity" and emphasized OpenAI's goal to use the technology to "accelerate defenders" and deliver "the transformational benefits of our cyber models to more organizations in Japan"2

2

4

4

.

Why AI-Powered Cyberattacks Demand AI Defenses

The launch addresses a fundamental shift in the threat landscape: attackers are already using AI to automate and scale their assaults, forcing defenders to adopt equally sophisticated AI-savvy approaches2

2

. The use of AI has caused the number of attacks to balloon exponentially and grow more complex, meaning defenses must become more versatile1

1

. The same AI models that probe systems for flaws can be turned around to find vulnerabilities first, giving defenders a fighting chance2

2

.

No pricing was announced for the service, though SoftBank offered free vulnerability assessments to all attendees at Tuesday's Tokyo presentation1

1

2

2

. The official release describes a measured, progressive outreach to "selected eligible companies" rather than an immediate deployment across all 3,000 target firms2

2

. The success of this AI-driven cybersecurity approach will ultimately depend on whether organizations can close the security gaps faster with AI assistance than they could without it, and whether critical vulnerabilities actually get patched before attackers exploit them.