As organizations accelerate their innovation in the cloud and their adoption of AI, securing AI workloads and identities has become critical. Misconfigurations, vulnerabilities and identity-based threats expose high-value assets to potential manipulation and exploitation. For AWS customers, advancing cloud security means establishing resilient guardrails that protect innovation without compromising speed.
Speed is critical as adversaries accelerate their operations. The CrowdStrike 2024 Global Threat Report revealed adversaries’ average breakout time — the time it takes to move from an initially compromised host to another host in a target environment — is down to just 62 minutes. Cloud intrusions jumped 75% between 2022 and 2023, driving the need to tackle common challenges such as poor visibility, unprotected workloads and delayed threat response.
All the while, organizations face new threats to the very workloads and AI models driving their innovation. Images and containers deployed with vulnerabilities and misconfigurations can put critical infrastructure at risk. Attackers are quick to exploit weaknesses to manipulate outcomes, steal proprietary algorithms or even repurpose infrastructure for malicious ends. The need for a robust, trusted and comprehensive cloud security solution has never been more urgent.
CrowdStrike is pleased to announce several new innovations and integrations with AWS to modernize cloud security. Read on to learn more about:
- Innovations made to protect AI workloads
- Integrations with AWS to secure code before deployment
- How CrowdStrike enhances cloud detection and response across AWS
- CrowdStrike named Partner of the Year by AWS in several categories
Protecting AI Workloads
As organizations race to adopt cloud and AI, security teams struggle to keep pace. Emerging technologies can be gold mines for adversaries, exposing new attack paths that can be used as footholds into critical infrastructure. CrowdStrike’s new capabilities help organizations secure every stage of the software development lifecycle — from build to runtime to posture management — to secure AI deployments and protect sensitive data on AWS, enabling them to innovate with peace of mind.
Extending Support to Amazon SageMaker
At Fal.Con 2024, we announced AI security posture management (AI-SPM) features to detect misconfigurations and risks across platforms like OpenAI, Amazon Bedrock and Vertex AI, helping prevent AI-specific threats such as model tampering or poisoning. We are now excited to announce additional support for AWS SageMaker, a fully managed machine learning (ML) service.
With Amazon SageMaker, data scientists and developers can quickly build, train and deploy ML models into a production-ready hosted environment. Now with CrowdStrike supporting Amazon SageMaker, organizations can build confidently as they innovate in the cloud.
Scanning AI Container Images
CrowdStrike Falcon® Cloud Security now monitors AI model runtime behavior to detect and respond to threats in real time. With our new release, Falcon Cloud Security extends its protection to the build phase, covering the entire lifecycle from build to runtime and posture management. Falcon Cloud Security scans for vulnerabilities and misconfigurations, enabling organizations to secure AI workloads before deployment. This approach helps organizations identify and mitigate risks early, preventing breaches and ensuring the secure deployment of AI workloads.