Datadog, the monitoring and security platform for cloud applications, today announced at its annual DASH conference its latest security product, Datadog Security Platform. These advanced capabilities are engineered to counter the ever changing threat environment stemming from the increased usage of AI in enterprise settings, providing holistic prevention from conception to consumption.
As AI systems gain autonomy and are embedded in mission-critical processes, they present new attack surfaces and vulnerabilities which are difficult to be detected by traditional security mechanisms. Aware of this emerging challenge, Datadog have considerably expanded its security platform to address the needs of securing AI-native applications and agentic workloads.
One prominent feature of the announcement is Bits AI Security Analyst, an entirely new intelligent agent built into Datadog Cloud SIEM. This AI agent triages security signals for security teams, conducts full-scale investigations into threats, and delivers animated and contextualized actionable recommendations to enable security teams to address breaches 53x faster.
It takes a huge load off the backs of human analysts, so they can concentrate on the threats that are most important.
In addition, Datadog unveiled Code Security, also generally available, a solution that applies AI to identify and rank vulnerabilities in custom code as well as in open-source libraries, among them libraries used in Large Language Model (LLM) integrations. This solution also features AI-based remediation advice to ease the burden of resolving complicated problems.
And, for preservation of AI model integrity itself, this week Datadog introduced improved LLM Observability. This tool tracks the behavior of AI models, and how it does, by providing essential toxicity checks on prompts and outputs to detect and prevent harmful or inadvertent outputs.
In addition, new LLM Isolation capabilities, which are in preview, aim at catching and stopping attackers from exploiting holes in LLMs and enforcing guardrails for secure production AI models.
In response for data privacy, Datadog also unveiled a Sensitive Data Scanner (which is in preview now for AWS S3 and RDS instances) to stop confidential data from inadvertently spilling into LLM training or inference data sets.
DASH 2025 highlights these new innovations and demonstrates Datadog’s support in helping businesses to adopt AI with confidence. Datadog’s added preventative and responsive capabilities, combined with its end-to-end runtime visibility, are enhancing the security of AI workloads at every layer of the stack.