SentinelOne launches AI security tools for product and security teams
SentinelOne released four new AI security products on Wednesday, addressing two distinct problems: securing AI agents that companies are deploying into operations, and using AI to speed up security investigations.
The releases include Prompt AI Agent Security, Prompt AI Red Teaming, general availability for Purple AI Auto Investigation, and data pipeline functions within Singularity AI SIEM. Together they reflect a shift in how security teams operate as AI tools move into production workflows.
Controlling AI agents in production
Prompt AI Agent Security provides visibility and control over AI agents and agentic workflows. It monitors agent activity in real time, assesses risk, and enforces policy across environments where those agents operate.
Preview features will let customers manage agent security posture and automatically respond to risky behaviour-such as an agent sending company data externally or escalating privileges across enterprise systems.
For teams building their own AI applications, Prompt AI Red Teaming tests for AI-specific vulnerabilities that conventional security tools miss. It simulates prompt injection attacks, jailbreaks, privilege escalation, and data poisoning.
Automating investigation workflows
Purple AI Auto Investigation is now generally available. The tool triggers a full investigation with a single action, gathering evidence across data sources, compiling attack timelines, and producing verdicts that can trigger automated remediation-all while remaining under analyst oversight.
SentinelOne said the tool reduces investigations that previously took hours or days to minutes or seconds. It integrates into the Singularity platform and is available to existing Purple AI Analyst customers without additional setup.
Purple AI has become central to SentinelOne's business. During the company's fourth-quarter earnings call for fiscal 2026, it reported that Purple AI was included in more than 50% of all licences sold that quarter-a record attach rate.
Filtering data before it enters the SIEM
Following its acquisition of Observo AI, SentinelOne is integrating AI-native data pipeline functions into Singularity AI SIEM. The system filters, enriches, and normalises data before it reaches the main platform.
SentinelOne says this approach can reduce data noise by up to 80% before ingestion, lowering infrastructure costs and improving analysis across third-party data sources. The move addresses a broader industry trend: vendors are competing on their ability to handle the growing volume of machine-generated security data earlier in the pipeline.
Why this matters for product teams
Security teams managing AI deployments now face two linked challenges: protecting AI tools themselves and using AI to manage increasingly complex security operations. AI for Product Development teams need to understand how their AI systems will be monitored and secured in production.
Market forecasts cited by SentinelOne point to rapid growth in AI-related security spending. Gartner projects AI cyber security spending will grow at a 73.9% compound annual growth rate from 2024 to 2029, making it one of the faster-growing segments of AI expenditure.
These releases also reflect the operational reality of AI Agents & Automation. As companies deploy autonomous agents into production, security teams need visibility into what those agents do and the ability to stop them before they cause harm.
Your membership also unlocks:
