Artificial Intelligence · Cybersecurity
Wazuh adds AI tools for SOC workflows
Wazuh described AI features for its cloud service and self-hosted deployments, including scheduled reports, local model use and a dashboard assistant built on third-party AI services.
- Wazuh said it offered the Wazuh AI Analyst on Wazuh Cloud and integrations with third-party AI providers.
- The company said the AI Analyst processed subscription data through Amazon Bedrock and Anthropic's Claude and sent scheduled email reports with indicators, endpoint counts, alert activity, active vulnerabilities and a posture summary.
- Wazuh said the subscription data was not shared with third parties or used to train models, and that the output should be checked against local policy before action was taken.
- For self-hosted deployments, Wazuh said Ollama could run Meta's Llama locally on the Wazuh server and expose a chatbot over archived logs.
- It also said Anthropic's Claude 3.5 Haiku could be used through Amazon Bedrock in the dashboard via OpenSearch Assistant.
Wazuh presented AI as a way to reduce repetitive work in security operations rather than replace analysts. It said its cloud service and self-hosted deployments could both make use of AI tools.
For Wazuh Cloud, the company said the AI Analyst handled subscription data through Amazon Bedrock and Anthropic's Claude. It said the service sent scheduled reports by email and also made them available in the cloud console.
Wazuh said those reports included key indicators, endpoint coverage, alert volume, active vulnerabilities and a summary of the security posture. It said the data was encrypted in transit, processed in isolation and not kept permanently.
The company also described a self-hosted option in which Ollama ran Meta's Llama locally on the Wazuh server. In that setup, a script prepared archived logs for a chatbot, and a separate integration exposed Claude 3.5 Haiku in the dashboard through OpenSearch Assistant.
Named in this story
Companies
- Wazuh
- promoted its AI-related security features
- Anthropic
- provided the Claude models used in the article's integrations
- Meta
- provided the open-source Llama model family
Products and systems
- Wazuh AI Analyst
- provided automated AI security reports for Wazuh Cloud
- Wazuh Cloud
- hosted the AI Analyst service and report console
- Amazon Bedrock
- hosted the model service used in Wazuh integrations
- Claude 3.5 Haiku
- was surfaced in the dashboard through an integration
- Ollama
- ran a local Llama model on the Wazuh server
- Llama 3
- was offered as a self-hosted model option
- OpenSearch Assistant
- presented Claude 3.5 Haiku inside the dashboard
How the source tells it
The source read as vendor promotion with a reassuring, hands-off AI framing.
- vendor boosterism the piece repeatedly presented Wazuh features and integrations as reasons to adopt the platform
- reassurance it stressed that the tools augment analysts, keep data private, and leave validation to humans
- novelty hype it cast AI as a defining shift and described the service as automated and requiring no manual setup