Cybersecurity threats are evolving faster than many organisations can handle, and to counter that, Google has recently introduced AI Threat Defence, a new cybersecurity platform. The platform is designed to continuously monitor systems, detect vulnerabilities in real time, and help businesses respond before attackers can gain access to sensitive data. With the introduction of AI Threat Defence, Google has now entered into direct competition with Anthropic’s Mythos and OpenAI’s Daybreak. These tools have been launched previously and have gained attraction because they scanned a bunch of code and found vulnerabilities overnight.
Attackers are using AI to scan networks, locate weak points, and plan cyberattacks within hours rather than days or weeks. Google believes businesses require security platforms that operate continuously, identify threats in real time, and respond before attackers can compromise sensitive systems or steal critical data. The company further added that the conventional cybersecurity systems are no longer sufficient to combat modern threats.
Francis deSouza, COO of Google Cloud and President of Security Products, said the platform is designed to help organisations prioritise the most urgent threats instead of overwhelming security teams with endless warnings and alerts.
Google’s AI Threat Defence platform combines the company’s Gemini AI models with cybersecurity technologies from Wiz, Mandiant, and CodeMender. The platform follows four main steps in its AI Threat Defense which include prepare, scan, remediate, and monitor.
Google AI Threat Defence starts by reducing unnecessary exposure of the companies by identifying internet-facing systems and risky access points. Following that the tool then scans all the applications, cloud systems, APIs, and identities to find weaknesses that attackers could exploit.
Once a flaw is found, CodeMender and Gemini can suggest fixes directly inside developer tools. The system can also generate tests to check whether patches work correctly before they are deployed. Google says this could reduce the time needed to fix vulnerabilities from weeks to minutes.
Also read: AI is going to bring huge economic changes, says OpenAI
Not only that, but the tool can also analyse software dependencies and automatically generate tests to verify whether security fixes work correctly before deployment.
Google says this automated approach could reduce vulnerability remediation timelines from weeks to just minutes, helping organisations respond faster to emerging threats while minimising operational risks.