Sysdig extends its AI Workload Security to Amazon Bedrock, SageMaker, and Q, addressing security challenges in cloud environments posed by AI workloads. The enhanced security solution offers real-time visibility, detects suspicious activity, and prioritises vulnerabilities in AI applications.
Sysdig Expands AI Workload Security to Amazon’s AI Services
By [Your Name]
PHILADELPHIA – At this year’s AWS re:Inforce conference, Sysdig announced the extension of its AI Workload Security to Amazon Bedrock, Amazon SageMaker, and Amazon Q. Sysdig, known for its real-time cloud security solutions, aims to address the growing security challenges posed by AI workloads in cloud environments.
As more organisations adapt AI technology, security remains a critical concern. AI workloads, which often contain sensitive training data, represent lucrative targets for cyber attackers. Sysdig’s AI Workload Security, part of their cloud-native application protection platform (CNAPP), enhances security teams’ capabilities by offering real-time visibility, identifying suspicious activity, and prioritizing potential vulnerabilities.
Amazon’s suite of AI services, including Bedrock, SageMaker, and Q, facilitate the development of generative AI applications by providing foundational models and seamless integration with AWS environments. According to AWS, over 10,000 organisations globally are leveraging these AI services. However, the adoption of generative AI introduces additional security risks. Research by the Sysdig Threat Research Team indicates that generative AI workloads have a 35% increased likelihood of becoming publicly exposed, compounding these security challenges.
The difference between on-premises and cloud environments in terms of speed, complexity, and dynamism is substantial, and AI exacerbates these security risks. Organisations are often required to detect, investigate, and respond to attacks within minutes. Sysdig, in collaboration with AWS, seeks to secure the rapid integration of AI into business processes by utilizing real-time detections and deep runtime visibility to counter these threats.
Sysdig’s enhanced AI Workload Security leverages real-time signals from AWS CloudTrail logs to mitigate incidents such as:
-
Reconnaissance activity: Detects attempts to discover and exploit AI services, giving security teams the ability to preempt malicious activities.
-
Data tampering: Identifies efforts to manipulate data, delete models, or disable logging, preserving the integrity of AI applications.
-
Public exposure: Highlights instances where AI applications are exposed to the internet, aiding teams in securing sensitive information.
“While there is a rush to integrate AI into software, understanding AI risk and implementing appropriate security measures is crucial,” remarked Loris Degioanni, CTO and Founder of Sysdig. He emphasised the collaboration between Sysdig and AWS to help customers safely leverage AI’s efficiency and speed.
Sysdig’s expertise in cloud threat detection, highlighted by their development of the open-source standard Falco, positions them to aid organisations in managing AI usage. Their platform integrates real-time AI Workload Security with unified risk findings, providing security teams with a consolidated view of risks and events. This consolidation facilitates more efficient prioritisation, investigation, and mitigation of active AI risks.
The announcement comes as Sysdig aims to support organisations in managing the complexities of cloud security without hindering innovation. At AWS re:Inforce, attendees can visit Sysdig’s booth for demonstrations of their AI Workload Security capabilities. The company continues to build on its reputation for delivering real-time protection and insights to combat cloud threats.
