SentinelOne targets AI risks across lifecycle 

SentinelOne has expanded its AI Security Platform with new data security posture management capabilities aimed at protecting enterprise data across the entire AI lifecycle, Security Brief Australia reports. 

The company said the move reflects the shift of AI models and agents from experimentation into production, where larger volumes of sensitive corporate data are now feeding training and inference pipelines. The new DSPM tools focus on cloud object storage and databases, helping security teams identify and control high-risk data before it enters AI systems, where removal can be difficult. SentinelOne also highlighted threats such as pipeline poisoning and malicious data ingestion, adding scanning controls at the data entry stage. DSPM integrates with the vendor's broader platform, which includes cloud security posture management, AI-specific posture controls, runtime workload protection, and safeguards for employee use of generative AI.

At runtime, protections extend into cloud workloads hosting AI models to address risks like lateral movement and unauthorized access. SentinelOne tied the approach to growing regulatory pressure, arguing that AI security requires continuous visibility across data, infrastructure, and operational environments rather than isolated controls.