Google launches AI cybersecurity platform to counter rising AI threats

The new Google Cloud platform brings together AI models and security tools to detect exploitable vulnerabilities, reduce false alerts and automate response actions as enterprises face rising risks from AI-enabled cyberattacks.

Google Cloud has unveiled a new artificial intelligence-driven cybersecurity platform, AI Threat Defense, aimed at helping organisations detect, prioritise and neutralise digital threats in an increasingly AI-powered attack landscape. The launch reflects growing industry concerns about AI tools being used to rapidly uncover software vulnerabilities at scale.

The company said the platform is designed as an always-on security layer that continuously monitors enterprise systems, identifies potential weaknesses and focuses attention on threats that can actually be exploited in real-world environments. By filtering out low-risk alerts, the system aims to reduce operational pressure on cybersecurity teams while improving response speed.

Rising AI-driven threats reshape cybersecurity priorities

The launch comes amid intensifying competition in AI security technologies. Recent months have seen several major developments, including Anthropic’s introduction of its cybersecurity-focused model Claude Mythos and OpenAI’s GPT-5.5-based Daybreak platform. These systems have highlighted both the potential and risks of AI in identifying vulnerabilities across complex software environments.

Google Cloud executives said modern security teams are struggling with the sheer volume of alerts generated by automated vulnerability scanning tools. Many flagged issues, they noted, may not be practically exploitable, leading to inefficiencies and delayed responses to genuine threats.

Francis deSouza, Chief Operating Officer of Google Cloud and President of Security Products, said the platform is designed to help organisations “actively predict attack paths, prioritise the most significant threats, and deploy verified fixes faster than adversaries can exploit them.”

AI-powered analysis and automated remediation

AI Threat Defense integrates multiple layers of intelligence, combining Google’s security capabilities with technologies from Wiz, Mandiant and CodeMender. The platform evaluates whether vulnerabilities are exposed to the internet, reachable through network configurations or realistically exploitable within an organisation’s environment.

To improve accuracy, the system uses a mix of AI models. Lightweight models handle continuous large-scale scanning, while advanced Gemini-based models are applied to high-risk assets requiring deeper analysis. The platform also simulates attack scenarios using AI agents to validate whether identified weaknesses can be exploited in practice.

Beyond detection, Google has introduced automated remediation features. AI agents can assist in rewriting outdated code into modern, memory-safe programming languages under human supervision. The system can also analyse dependencies, generate tests and validate fixes before deployment, reducing the risk of introducing new vulnerabilities.

In addition, the platform supports continuous monitoring and automated response workflows, enabling AI agents to investigate suspicious behaviour and assist security teams in real time.

Google said ecosystem partners including Accenture, Deloitte, PwC, Netenrich and TENEX.AI will help organisations deploy and integrate the solution into existing security frameworks.

The company emphasised that as AI-driven attacks evolve rapidly, traditional manual vulnerability management is no longer sufficient, making automated, intelligence-led defence systems essential for modern cybersecurity strategies.