Google takes on Anthropic and OpenAI with new cybersecurity platform
What's the story
Google has launched a new AI-powered cybersecurity platform called AI Threat Defense. The tool is designed to tackle real-world threats and provide automated fixes. It comes as a response to the rising concerns over powerful AI tools that could potentially compromise any software or operating system. The launch of this platform follows Anthropic's introduction of its cybersecurity-focused AI model, Claude Mythos, in April and OpenAI's launch of the Daybreak platform based on GPT-5.5 earlier this month.
Platform features
How AI threat defense works
Unlike its predecessors, AI Threat Defense focuses on real-world threats and automated fixes. The platform aims to reduce alert fatigue by prioritizing exploitable vulnerabilities and verified solutions. "AI Threat Defense helps organizations actively predict attack paths, prioritize the most significant threats, and deploy verified fixes faster than adversaries can exploit them," wrote Francis deSouza, COO of Google Cloud and President of Security Products, in a blog post.
Alert management
Addressing alert fatigue in cybersecurity
One of the major challenges with existing AI cybersecurity systems is that they inundate security teams with thousands of AI-generated alerts. While Anthropic Mythos and OpenAI Daybreak can quickly identify a large number of vulnerabilities, not all are actually harmful in real-world scenarios. Google AI Threat Defense tackles this issue by integrating code scanning with the cloud-security platform Wiz to determine if a vulnerability is reachable from the internet or exposed through live network configurations.
Threat assessment
Prioritized fixes for security teams
If a high-severity flaw exists in the code but is completely isolated, the system lowers its priority. This way, developers and security teams can focus only on active threats that attackers could realistically exploit. "Unlike other model providers that simply hand security teams a massive, unprioritised list of AI-generated alerts, we deliver prioritized fixes to accelerate remediation and secure the Defender's Advantage," wrote deSouza in his blog post.
System capabilities
Continuous deep-dive scans for threat detection
Google AI Threat Defense continuously performs deep-dive scans across millions of lines of corporate code. The system employs lighter, more cost-effective AI models for broad continuous monitoring, while Gemini frontier models are deployed for high-risk assets and complex exploitability analysis. This two-pronged approach ensures comprehensive coverage against potential cyber threats.
Fix implementation
Autonomous AI agents rewrite legacy code
Unlike OpenAI's system, which mainly suggests patches, Google uses autonomous AI agents under human supervision to actively rewrite legacy code into modern and memory-safe programming languages. The platform also analyzes dependencies and automatically generates tests to verify whether a fix actually works before deployment. This proactive approach sets Google apart from other model providers in the cybersecurity space.