As the time from vulnerability discovery to exploitation has fallen to 1.6 days, Google has presented an AI-powered defense design. It explained an operating model that uses multiple AI models selectively and is based on prioritizing tasks.
Google SecOps is said to make response 30 times faster with the introduction of AI, meaning that 10 analysts could handle the workload of 300 people at the same time. The report said AI is expected to accelerate tasks such as repetitive alert response and incident processing.
As part of its AI operating design, Google explained a policy of using multiple models selectively for each task. Rather than investigating all assets indiscriminately, it is considered essential to prioritize asset selection and the order in which vulnerabilities are remediated.
An IDC Japan survey reflecting security awareness among Japanese companies found that ransomware was the greatest threat, against a backdrop of a series of major incidents in 2025.
Concerns including entering confidential information into external AI services and vulnerabilities in AI-generated code were reported to be of particularly high interest among larger companies.