SoftBank Group is working with OpenAI on a cybersecurity offering that uses advanced AI to detect and respond to increasingly sophisticated AI-based attacks.<\/p>\n
The product, called Patching as a Service, is designed to support enterprise-wide cybersecurity vulnerability assessments and remediation planning. It is initially targeted at Japan\u2019s largest \u00a0companies, including critical infrastructure operators such as power systems and mass transportation providers.<\/p>\n
The two companies<\/a> previously set up a 50:50 joint venture, SB OAI Japan, to develop and market AI services in Japan. Ahead of the launch, SoftBank conducted a large-scale internal vulnerability assessment of its own systems using OpenAI\u2019s cybersecurity tools.<\/p>\n SoftBank plans to make the program available to selected companies in Japan and has started accepting applications for vulnerability assessments.<\/p>\n A New Phase of Cyber Risk<\/p>\n The rise of AI-driven attacks\u2014from deepfakes to targeted phishing campaigns\u2014has added complexity to enterprise cybersecurity. Research from HYPR found that a growing share of organizations now view generative artificial intelligence as a major security concern<\/a>.<\/p>\n In response, many enterprises are exploring the use of large language models (LLMs) to support tasks such as vulnerability discovery and remediation planning. The Softbank approach integrates LLM outputs into structured workflows, including ticketing systems and patch validation processes.<\/p>\n The Battle at the Enterprise Level<\/p>\n AI is especially valuable for cybersecurity at the enterprise level. Reliance on AI can dramatically decrease the amount of time cyber teams spend looking for gaps or identifying questions that challenge their corporate-culture assumptions.<\/p>\n \u201cCyber teams can enhance their workflows and identify cybersecurity gaps by testing and challenging their assumptions through AI,\u201d said Tracy Goldberg<\/a>, Director of Cybersecurity at Javelin Strategy & Research. \u201cFor instance, cyber teams can rely on AI to help determine which exploits attackers might use to gain access to their systems and networks, based on the size of their organization, public profile, and other known data points that can be input into the AI platform.\u201d<\/p>\n