Red Hat, a global leader in open source solutions, has announced the acquisition of Chatterbox Labs, a leading company in AI model security and protection system development for generative AI. With this move, Red Hat adds key “AI security applied” features to its Red Hat AI portfolio, strengthening its goal of delivering an open, comprehensive enterprise AI platform optimized for hybrid cloud environments.
The acquisition occurs amid rapid evolution of Red Hat AI, following the launch of Red Hat AI Inference Server and the introduction of Red Hat AI 3. Organizations across multiple sectors and regions are already adopting these solutions to accelerate generative, predictive, and agent-based AI projects. As companies transition from experimental to production stages, they face the challenge of deploying models that are not only advanced but also reliable, verifiable, and secure. In this context, security and protection mechanisms are now essential elements of modern MLOps. This focus on trust and safety underscores Red Hat and IBM’s commitment to responsible AI adoption in hybrid cloud environments. Integrating Chatterbox Labs’ technology will enable a unified platform with security built-in from the ground up, expanding Red Hat’s capacity to run AI workloads in production with any model, accelerator, or environment.
Addressing Unintended AI Consequences
Founded in 2011, Chatterbox Labs provides critical technology and expertise in AI security and transparency. Its work on quantitative AI risk assessment has been recognized by independent global expert groups and policymakers, and this acquisition brings advanced machine learning technology to Red Hat.
Chatterbox Labs offers automated, customizable AI safety testing and protection capabilities, delivering objective risk metrics that executives need to approve deploying AI in production. Its technology provides a model-agnostic, robust approach to validating data and models through:
- AIMI for Generative AI: Provides independent quantitative risk metrics for large language models (LLMs).
- AIMI for Predictive AI: Validates any AI architecture across key pillars, including robustness, fairness, and explainability.
- Protection Mechanisms: Detects and corrects unsafe, toxic, or biased prompts before models go into production.
Securing the Next Generation of AI Workloads
This acquisition aligns with Red Hat’s vision to support diverse models and deployment environments in hybrid cloud setups. It also complements future-oriented capabilities introduced in Red Hat AI 3, specifically for agent AI and the Model Context Protocol (MCP). As enterprises adopt agent-based AI, reliable and secure models become even more critical due to the complex and autonomous nature of AI agents and their potential impact on core business systems.
Chatterbox Labs has conducted research on holistic agent security, including monitoring agent responses and detecting server MCP triggers. This work aligns with Red Hat’s roadmap for supporting Llama Stack and MCP, positioning Red Hat to secure the next generation of smart, automated workloads with a foundation of trust and enterprise readiness. By combining Red Hat’s MLOps capabilities with Chatterbox Labs’ protection mechanisms, Red Hat will enable organizations to operationalize their AI investments with greater confidence.