Red Hat announces the acquisition of Chatterbox Labs to add AI security and guardrails to the Red Hat AI portfolio.
Red Hat has announced the acquisition of Chatterbox Labs, a company focused on model-agnostic AI security and guardrails for generative AI. The technology should help organizations bring AI models to production in hybrid cloud environments in a demonstrably safer and more reliable manner.
AI Security
Organizations today face the challenge of deploying models that are not only powerful, but also reliable and secure. Security and guardrail functionality is therefore becoming a basic requirement for modern machine learning operations (MLOps).
Chatterbox Labs delivers automated, customized AI security and safety testing with the factual risk statistics that decision-makers need to approve AI applications for production. The technology offers a model-agnostic approach to validating data and models through:
- AIMI for gen AI, which provides independent, quantitative risk metrics for large language models (LLM’s);
- AIMI for predictive AI, which tests every AI architecture against key pillars such as robustness, fairness, and explainability;
- Guardrails that identify and mitigate unsafe or biased prompts before models are put into production.
Agentic AI and MCP
Red Hat links the acquisition to the growth of agentic AI and support for Model Context Protocol (MCP). The company points out that AI agents are becoming more autonomous and can thus have more impact on critical systems. According to Red Hat, Chatterbox Labs conducted research into agent security, including monitoring responses and detecting actions on MCP servers triggered by agents.
read also
Red Hat Injects more AI and Security into Red Hat Enterprise Linux
Red Hat wants to combine the guardrails and tests with its MLOps capabilities. The goal is to give organizations more control over AI risks in deployments with different models, accelerators, and environments, without being dependent on one specific model type.